In the age of tech and remote work and air travel, what does it mean to call a place home? There is an alternative to moving to a tech hotspot like Silicon Valley or New York, however less glamorous: getting involved in your local community through politics and policy.

We explore this path through a conversation among Eliza Steffen, Charles Yang, and Ashwin Ramaswami. We examine our shared interest in technology and government, how this led us to work at the local level and contribute back to our communities, and how faith and religion has shaped this mindset.

Subscribe now

The Future is Local

By Eliza Steffen, Charles Yang, and Ashwin Ramaswami
Conversation moderated by Jacob Kuppermann

How did you become friends?

Eliza: Ashwin and I first met on this trip called Perspectives back in summer 2018, which is a ten-day trip run by Stanford Hillel to Israel and the West Bank for non-Jewish student leaders, but I don't think we talked very much on the trip. We ended up getting reacquainted last April and we both met Charles!

Charles: Right, the three of us met in Washington, D.C. this past April at a birthday party, where we basically stood in a corner for two or three hours just talking. I think my first impression of Eliza was that she described herself as someone who “grew up Unitarian Universalist but now wants to convert to Judaism,” which was an unusual combination of things to me.

Eliza: Well, it worked because we're still here.

Ashwin: Yeah! We all talked about religion, pretty much right off the bat. I talked about the work I’ve been doing with Dharmic Life at Georgetown, where we’ve been building and sustaining a chaplaincy for students from Hindu, Buddhist, Jain, Sikh, and other Dharmic traditions. Somebody pointed me to Charles, telling me, “Charles is a minister, you should talk to him.”

Charles: To be clear, I don’t know why people introduce me like that! I am currently in seminary school studying theology, but am by no means a minister.

Was there anything unexpected that came up when the three of you met for the first time?

Ashwin: I remember learning about YIMBY from you, Charles. So thank you for introducing a new word. I had heard of NIMBY, but not YIMBY.2

Charles: Glad I could share the good news with you, Ashwin. 

I know you all have had experiences on the more national policy level, but how did all three of you come to care about local issues? 

Eliza: I grew up Unitarian Universalist (UU), which is a religious movement descended from Protestant Christianity that was founded when the American Unitarian Association and the Universalist Church of America merged in the early 1960s. In practice, UUs treat civic engagement like a religious obligation. My congregation provided plenty of opportunities to explore that. I grew up very involved in reproductive rights advocacy at the local and state level, often through my church, including both Unitarian and interfaith work. So it was like very much the beginning of my development: my political motivations, goals, and community were very intertwined with my religious community.

I was a little more nationally focused in college and more recently had the opportunity to do some place-based philanthropy work on federal investments (BIL/CHIPS/IRA) in the Midwest.3 It’s been fascinating to see the parallels between that and Ashwin and Charles’ experience as they go back to their hometowns in a more personal capacity—to see what issues are the most serious, versus the most tractable, versus which ones are getting the most attention.

Charles: Yeah, I was reflecting on that in the new year because it's been a very unexpected path for me. Like most college students in the Bay Area, I came in with my own unique major and dreams (in my case, materials science and engineering) and exited as a computer science graduate, with a focus on AI. I was always a “STEM kid” growing up, but the Bay Area is a great example of how so many problems in America, e.g. housing, infrastructure, and transportation, are not ones of technology but of politics and people. At Berkeley, I, like every Berkeley student, experienced very viscerally the power that a local city council has over someone’s living conditions and what kind of housing is (or is not) available.

When I came back home after graduating, I got to know the local county Democrats club and ended up seriously considering running for state legislature, motivated both by my commitment to my hometown and my exploratory journey into policy my senior year of college. Ultimately I decided against it and decided to go into federal climate policy full-time (that story is one for another time!). But I did start a YIMBY group and a local electric bike library and non-profit e-bike rebate program.4 Although I ended up not running for office, I’d like to think I ended up just convincing Ashwin to do it, which is probably more effective anyways.

But Ashwin, I know we're a bit similar in that we both spent 18 years in the exact places we were born in but didn't actively consider local politics until later in life—tell us your story!

Ashwin: That's right. When I grew up in Johns Creek, Georgia, I didn't actively follow local politics very much; it wasn’t much on my radar. But there were a few instances that left a big impact on me.

One is that I grew up being part of a Hindu spiritual and religious organization called Chinmaya Mission. We wanted to build a community center in Forsyth County. And we had some problems and faced opposition around getting permits. We had lawyers and advocates from our end go to city council meetings, and we all showed up to protest and tell elected officials, “Hey, we want this thing to be built.” That was one big moment.

And the other moment was when I went to a charter school, Fulton Science Academy, for sixth and seventh grade. When I was in seventh grade, the charter school got shut down by the Fulton County Board of Education. So students and parents showed up at the school board meetings to protest. I remember sending cold emails to all of the various Board of Education people telling them, “Please don't shut our school down.” Though we were unsuccessful, that's probably one of my first earliest political communications that I did manifest.

One thing we realized is that, for our hometown community, it felt like we weren't really empowered to help make a difference. Rather, other people were making decisions for us. 

How has religion in particular been important for all of you as a motivating factor behind your life trajectories?

Ashwin: Religion has been important because it helped form my worldview, in a way that other worldviews didn't really give me. For example, science is helpful, but doesn’t answer more fundamental questions, like, who am I? What am I here for? I needed a framework to introspect and think about these questions. In my case, I was given that framework by the Vedanta philosophy from Hinduism. In ancient India, there isn’t the same distinction we have today between religion and philosophy: it’s just a way of understanding the world.

Charles: Yeah, me too. Especially now that I'm older looking back, religious spaces are, at their best, one of the only places where as a 12 year old you're grappling with very profound questions of “what does it mean to be good?” and “how do you live your life in relation to death?” Those are deep questions to be dropping on teenagers. But that’s at least one benefit that I have gained from my faith. The other thing is that my faith is very much tied to my sense of home. A lot of the community that I grew up with is also the same community that is from my church in Richmond. And so many of my values are manifested and anchored in my hometown.

And I think that’s why, for reasons I’m still untangling for myself, I feel drawn to public service, and why leaving a startup and going into the federal government was appealing. Pursuing a startup, with the buzz of “disruption” and the genesis-like claims of creation (and sometimes, eschatological echos of destruction) stands in contrast to finding a vocation where I can serve the people and communities around me.

Ashwin: I agree. I think that's also why I went to work for the government. There’s some aspect of public service being linked to the experience of grappling with questions about life and death when you’re 12. Those questions always stay with you. You can't stay happy just by working at a tech company and making ads.

Charles: The Sunday school to federal government pipeline is very real!

Eliza: I was also very involved in youth leadership, where I was responsible for large groups of my peers, from when I was fifteen or sixteen. And Unitarians tend to love principles but not necessarily rules, so I had a lot of agency and was essentially tasked with shaping my peers—in terms of their religious development, but also their opportunities for engagement in the public sphere. That responsibility definitely shaped me early on and forced me to think deeply about things like what my obligations were to different communities, and how to resolve conflict. 

Then, early on in college, a friend invited me to an event at Stanford Hillel, and then I just kept going back. I did a textual study program early on and learned about the idea of tikkun olam or “repairing the world” and Talmudic debate, which felt like it mirrored the Unitarian principles I grew up with. Getting involved in the Jewish community in college was definitely an investigation into what my religious values, rituals, and of course, actual belief systems are. And in that learning process and self-interrogation, I saw how a strong obligation to public service exists across religious communities, especially when I led an alternative spring break class and trip on religious pluralism with my friend, who was raised Reform Jewish.

Charles: Yeah I think unlike Eliza, there was always this bifurcation of the public and the private life, with the public life in school and then the private life and church, which I never had any issue with. But it was an interesting dichotomy that I'm still bridging in my social relationships today.

Ashwin: Here’s another example of that bridging. I wrote a piece while a senior at Stanford as my senior column at the Stanford Daily. In my public life, like at school, I would be a bit quieter. But in my private life, I was much more talkative. I would talk in an Indian accent in one situation, then change my accent in another situation, code switching.

But I wrote about how, after going to college at Stanford, those two things got combined, and it changed. I realized that these two parts of your life are not actually separate. They're connected. I had always felt like on the outside, I had to put on a mask or fit in with other people. And then I started to lose these inhibitions and realized that I can just do whatever I think is right—that's what changed for me.

All three of you have in your own way found approaches to getting involved in civic life that are not quite the standard path. Do you have any advice for people who might be looking at this interview, seeing the paths you’ve followed, and wanting to do the same?

Charles: It took a really long time for me. I was working remotely for an AI startup for a year and a half. And so the journey we're talking about took place over that time, when I was thinking about what I valued and how I wanted to spend my time. It was then that I realized the whole traditional trajectory I was on, starting from when I was a CS major at Berkeley, wasn't actually what I was actually interested in or wanted to do. Stepping out of that college bubble was really helpful.

It also helped that I was able to do a lot of things on the side—a lot of this advocacy work is something you can do as a volunteer and just explore. I also owe a great debt to a few friends who made the jump before I did and showed me that there was this other world, this other mode of agency, that I would not have thought of if not for them.

But while I spent time working and then readjusting my course, Ashwin jumped into running for office straight from law school! How did you think about that decision?

Ashwin: Even at Stanford, one thing I did was taking lots of other classes outside of my subject area. In fact, I took some creative liberties to count all my classes towards my graduation requirements! But that really helped me because I took a lot of the liberal arts, which a normal CS major might not have taken, but gave me a wider perspective.

I had technical experience, like working in the tech sector and startups. I loved, and still love, to code, but I really wanted to do something that would help other folks and where I was most needed. That's why I started maintaining open source software, and later joined the federal government at CISA (the Cybersecurity and Infrastructure Security Agency). It was an area where I felt much less replaceable. When I was an intern at Google, I was only one among thousands of engineers chipping away at a tiny and likely inconsequential piece of code. But when maintaining a popular open source software library, or working at the government, I was really doing something that was irreplaceable.

Eliza: Wow. Unlike Ashwin and Charles, I do not know how to code.

Ashwin: My time at the government was only possible because of the pandemic, as I was able to be enrolled at Stanford, home in Johns Creek, working remotely at the federal government, and also working remotely at various startups, which could not have happened otherwise.

After the rollercoaster that was the 2020 election in November—and finding myself at the forefront of our nation’s election security efforts—I decided to dedicate myself to a career of public service. I had worked with lawyers at CISA and believed a law degree was a great way to get additional context on how our systems, laws, and government work. I studied for the LSAT in December, took the LSAT in January, and applied to Georgetown in February. Through my work in law and policy, I continued to see first-hand the impact that talented technologists could have in government and public policy, in topics ranging from open source software to AI safety to false news.

But I soon realized that I could contribute even more at the state level, which often has less resources yet more opportunities for change than the federal government. This is why I worked at the GA Attorney General’s Consumer Protection Division, and recently decided to run for state senate. I want to give back and be a voice to focus on the issues that really matter to my communities: fully funding our education, increasing economic opportunity, reducing prices and improving access to healthcare, and public safety. And restoring integrity and expertise to our system against an incumbent who was indicted for trying to overturn our election results.

Eliza: I now work in philanthropy, at a donor-advising firm that works with a lot of large technology funders and with the California state government. While I was in college, I was more pessimistic about the tech world, and became much more positive once I moved to D.C.—due to meeting really thoughtful, public-service minded technologists like Charles and Ashwin. And it’s been fascinating to see the ripple effects of what the tech world has changed in the past ten years, especially how the rise in funders from tech—both billionaires and smaller donors through conveners like the First Principles Forum—has shaped the philanthropic world. There’s more funding for higher risk initiatives and approaches.

There’s an increased interest in funding democracy and governance work, which is what most of my background is in. More specifically, research has shown an increase in funding for tech-related sub-issues like misinformation, disinformation, and media policy, though it’s not clear how much of that new attention is coming from tech industry donors. There’s also some adjustment—every potentially effective intervention in protecting U.S. democracy can be measured with an RCT or similar measurement. Overall, I think philanthropy is headed in a positive direction.

Ashwin: One last bit about advice. A lot of people advised me not to go to law school. Many lawyers mentioned that they regretted it, because being a lawyer requires long hours or was unfulfilling; but one mentor told me law school is amazing and that I should do it. And I listened to his advice. And a similar thing happened with getting involved with politics!

So my advice would be if you are talking to a lot of people and they give you their advice, it only takes one person to tell you that a certain path is good. You can just listen to them or just listen to yourself. It’s ultimately your decision!

Eliza: Ooh, radical advice giving.

Share

Eliza Steffen works at the intersection of philanthropy, policy, and politics at Freedman Consulting. She previously served as a Tom Ford Fellow in Philanthropy on the Civic Engagement and Government team at the Ford Foundation. She has attended Stanford University and is from Columbus, Ohio. She was raised in a large Unitarian Universalist congregation and now broadly practices Reform Judaism.

Charles Yang is an Energy Supply Chain Fellow at the Department of Energy (DOE). He previously worked at an artificial intelligence (AI) hardware startup and did AI for Science research. He also started a Yes In My BackYard (YIMBY) housing advocacy chapter in his hometown of Richmond, Virginia where he was born and raised. He was raised in a small Chinese-American Baptist church and is currently in seminary studying Christian theology part-time.

Ashwin Ramaswami is an editor of Reboot and is currently running for Georgia state senate in his hometown of Johns Creek, Georgia. He has a background in technology law and policy; he previously worked on election security at the Cybersecurity and Infrastructure Security Agency. He has a Hindu religious background and is interested in Sanskrit and Vedanta philosophy. He has attended Stanford University and Georgetown University Law Center.

Reboot publishes essays on tech, humanity, and power every week. If you want to keep up with the community, subscribe below ⚡️

Subscribe now

microdoses

• Did you know that Rep. Maxwell Frost, first Gen Z member of Congress, is also a YIMBY?

• One of the first reactions Charles had when he met Eliza: “I have no idea what's going to come out of your mouth next. These are so many words and phrases that I never expected to hear in the same sentence. It's kind of like an intricate chef prepared dish. There's constantly a new flavor.” 

• Another big influence that showed me the importance of working in state and local government was helping start the Judicial Innovation Fellowship! - Ashwin

• Forget the Federal Reserve, YIMBYs are the real inflation fighters

• Tweet:

  

closing note

Do reach out to the authors via their linked Twitter accounts if you’re interested in following up or pursuing related paths!

Returning to our roots,

Ashwin, Charles, Eliza, & the Reboot team

Get your copy of Kernel!

Charles Broskoski is co-founder and CEO of Are.na and is based in Hudson, NY. He is currently working on Are.na full-time alongside a small team. 

Meghna Rao is an editor and writer from Queens, thinking about diaspora and technology.

This interview has been edited for clarity and length. 

Taking Business Personally

Interview and intro by Meghna Rao

Here’s a story you might find familiar: you sign up for a startup service one day. Let’s say, for the sake of this example, that it’s a social media network that lets you upload and share photos with your friends. You find this fun. You take photos of silly moments, moving moments, beautiful moments. The images spawn a culture unique to the platform. Memes become self-referential. More friends join, and so do extended family, celebrities, your favorite magazine. 

One day, a few years down the line, you look up and realize that it’s no longer the same service you signed up for. Your feed shows more brands than family, and you learn about your newborn cousin two weeks late. The people you’re following are posting sponsored content, or they’re not posting at all. You’ve been given an endless scroll, so you find yourself falling through hours of recommended videos of strange recipes and struggling to exit.

It’s this feeling that Are.na, a platform for connecting ideas, is trying to avoid — and it’s doing so by eschewing traditional expectations of startup growth. 

Are.na is 12 years old, just a few months younger than Instagram. I found my way to the platform on a friend’s recommendation when I needed a place to put all the weird and varied media that influenced my internet brain — TikToks, screenshots, PDFs. You can use Are.na to link all of your weird online stuff, my friend told me. You can organize it into channels and browse through other people’s weird online stuff. 

In some ways, Are.na felt familiar to other startups I’d encountered — easy to use, obvious how to sign up and unsubscribe, transparent about how they’re building their product. But there were other parts that were unfamiliar. Namely, that there are no ads on Are.na, that there are still active users from a decade plus, and that the startup limits its expansion plans to grow mostly by word-of-mouth.

It’s a small startup that is growing at a steady pace. It currently has 25,000 monthly active users (MAU), 0.00125 percent of Instagram’s reported 2 billion MAU as of October 2022. Around half of Are.na’s active members pay for premium services. As of July 2023, those subscriptions come out to around $70,000 monthly recurring revenue. They hope to reach $75,000 by the end of July, and update this goal each month.

A little over half the money goes to payroll (two full-time and four part-time workers). The remaining goes to everything else needed to run a business — servers, website hosting, software, taxes and other fees. They’ve never raised formal venture capital, but in 2018 participated in a ~$270K equity crowdfunding round that allowed people to invest to own a small stake of Are.na. Each new thing the team builds, co-founder Charles Broskoski tells me, is to ensure that the people using Are.na continue to like the product and stick around — that means monthly updates like fixing search bugs, editing design details, and adding features to the mobile app. 

This is what Are.na believes in: Superfast is not the only pace for growth. It is possible to sustain and grow a digital business without throwing venture capital (VC) money at it. You can still build something lasting on the internet that looks the way you want it to. Eventually, Broskoski hopes Are.na will become the “next ​​Nishiyama Onsen Keiunkan,” a hot spring spa founded in Hayakawa, Japan in 705 A.D. — a sustainable business that is around for the long-haul.

🌱 The origins of Are.na

Meghna Rao: You’ve answered this question in other places, but I think it’s important to ask again. How did you start Are.na?

Charles Broskoski: I think it’s important to say first that the throughline of this story — my thesis — is that more people should be starting businesses on the internet. It’s doable within certain parameters. And I want people to know that what they see when looking at Are.na today is just the result of a series of situational decisions we made along the way.

I have a background in art. Specifically, art on the internet. I had been hanging out on the internet on a platform called del.icio.us, which was the first social bookmarking website.

You could save links. You see a timeline of your bookmarks, and get a sense of where you were ten days ago, and what it led into next. You could see who else had saved the things that you thought only you were interested in, and you could also see the trajectory of what they were thinking about.

Similar to Are.na in some ways. But unlike Are.na, there wasn’t a social, communication component, right? 

Exactly. Like Are.na, Del.ici.ous was first and foremost a personal utility. You could tell if you would get along with someone or not based on what their perspective was pointed towards. And I made friends with someone named John Michael Boling. There wasn’t really a direct message function but we’d follow each other’s work. 

John was also making art at the time. We were both part of a scene. But then Yahoo bought del.icio.us, and there was no sense of what to do next. It wasn’t a political position, it was just like — OK, I guess we’re done here. We were like, they’ll fuck it up now. And they did; they did all of these things that made it unusable.

Everyone left. John Michael was badgering me to start my own version of del.ici.ous, and he wasn’t a programmer but he’d show me all these mockups and screenshots. He was looking everywhere for a short, interesting URL to purchase, and he found Are.na. 

Then there was me — I was starting to feel weird about doing art commercially. I’d recently had a solo show, and it didn’t make me feel good. When I got back, John Michael had met this person he knew through his sister, J. Stuart Moore. He was the dictionary definition of an eccentric millionaire. He’d started a company in the ’90s, made a ton of money, and retired. They got to talking. 

We were in agreement on some things. Like, there was so much stuff on the internet, we loved learning and engaging with work online, and we wanted to figure out a way to harness that more directly. We all went along with it. We hired three of our friends, all artists. Damon Zucconi, who is an artist and brilliant engineer and Dena Yago, also a brilliant artist who worked on K-HOLE. 

So Are.na really got its start because of J. Stuart Moore’s investment and his belief in your work. 

It wasn’t even an investment. We were employees. I was paid a salary. I also had that odd experience of making more money than I’d ever had before, after being primarily an artist and making zero money.

So many artists experience that surge when they get a full-time tech job.

Yeah, and Stuart hired all artists. We had very specific ideas about what we wanted. Our bar was basically — can we get all of our snobby, intelligent friends to use it without it being a favor to us? Can we make something actually useful for them?

We could, and this was a huge win but Stuart didn’t see it like that. Still, we were like — no, this means something. 

Stuart wanted there to only be certain types of acceptable content. Not frivolous memes. He thought the platform would be better if we controlled what went into it, but we were the opposite. As people who were embedded in this world, we knew — you never know what will lead you into something good and serious. Sometimes you have to take weird, windy paths. We thought it would be better if we just attracted smart people to it.

Our approaches started to diverge. And to Stuart’s merit, instead of firing us, he said — you guys care a lot about this. So I’ll give you the company.

Stuart changed my life with that decision. But suddenly I had a startup. None of us ever wanted that for ourselves. I’ve had a website since I was 13, and Damon and I are self-taught programmers, but prior to this we didn’t have any desire to be entrepreneurial. 

We languished for maybe a year. We kept the servers on. Our friends were using it, some really actively. But it was extremely small-scale. There was no business model. 

Everyone else was losing interest — except for me and Chris Sherron, who did all of the design. Together, we formed a new company and transferred all the intellectual property. We formed it as a Delaware C-Corp. I got a job because I wanted to get better at programming. And I convinced Daniel Pianetti, who I met through Are.na and who had all these ideas for what he wanted for the platform, to work with us. 

Were you nervous at all? Scared that you had a company on your hands now, and that you had no experience? 

Yeah, I was. The adjustment took a long time. But Are.na will be 12 years old in August. I wouldn’t be running a business if it weren’t for Are.na. It’s a very personal business, meaning that it is something I want to see in the world; it is a tool that I would personally be devastated if it were to not exist.

I think we have so many examples of assholes starting businesses, and if they’re the majority of examples, we’re only going to have shitty businesses. I want to see people with very specific perspectives and opinions on how to manifest things into the world. More people should take business personally.

💸 What wouldn’t we be embarrassed by?

You’ve only ever crowdfunded. Were you tempted to take traditional VC money and grow faster? 

Yeah, when we formed a new C-corp it was in the middle of the [early 2010s] startup boom, and so many people were raising money on nothing. I even figured that we could try raising a little. We tried to apply for accelerators, but nothing ever felt right. It all felt off brand, you know?

We got a grant from the Knight Foundation in 2016. It was $35,000, which wasn’t much, but it was wind in our sails. And that money felt on brand.

We started to ask ourselves — what wouldn’t we be embarrassed by? That became our guiding star. 

So it took a series of small yeses and nos to help shape Are.na into what it is today. 

Exactly. The set of decisions that helped us develop what we wanted to do. We would talk to people — VCs, accelerators — and I think it never worked out because we didn’t want to do anything embarrassing. We didn’t want to feel bad or make a decision that felt wrong. 

When did you cross the threshold into being comfortable enough with Are.na to sustain yourself?

So recently. I stopped doing freelance work to offset my salary two years ago.

It was hard getting to the place where we are now, where we feel comfortable. And I’m really happy we did. It feels so much better than if we were on the hook for something, or working with someone we didn’t like. 

How do you measure Are.na’s success since you’re not following the traditional startup growth playbook? 

We’re pretty in-touch with people. If someone writes in to help, it’s either me or Daniel who’s answering the email. We have a Discord where people tell us about what’s going wrong. I run our social media accounts. All of this makes our surface area of being connected to people very large, and people can express their satisfaction or dissatisfaction easily. So we always have a gut feeling about whether things are going right or wrong. But also, the only metric we really pay attention to is revenue. For us, it’s the highest bar. It sounds crass, but if someone is willing to pay for Are.na, that’s huge. Sometimes, we also track page views. No demographic or location. We don’t have time for all that.

📈 The goal is to be good

But certain things — personalized feeds, maybe — help manage larger platforms as they grow, right? 

It’s funny, I’ve never bought the idea that personalized recommendations are a response to growth. I feel like that’s an excuse to build an endless (addictive) content stream that’s financed by ads. Because we grow at a slower pace than most social platforms, and because we are fairly close and available for feedback to the people who use Are.na the most, we can do this more organically, for now. 

End-experience wise, I really can’t imagine a world where something like an algorithmic feed is the best response for many more people using Are.na, because it really kind of defeats the purpose of the whole thing. My feeling is that algorithmic feeds are just disrespectful; they disregard all the nuance in taste a person has. For applications with heavy personalized recommendations (I’m thinking about some streaming services, or even TikTok), I would personally be so much happier and interested with a truly random output.

To date, the biggest product change we’ve done with growth is to make sure the infrastructure can perform with increasing load. That was a large amount of work. We started it because people were complaining about slow load times, and we experienced it ourselves, as users. 

Do you ever worry that a much faster competitor will come in and make the same thing and beat you to it?

No, not at all. The reason the quality has remained so high on Are.na is because we grow slowly, and we have the patience to do that. And that’s why I think we can be in this for a long time: the primary goal is not to to make a shitload of money. 

Having a strong point of view about this type of thing is a meaningful differentiator. A lot of other platforms are diluted; their ethics do not translate to their product choices. People who discover Are.na are like, wow, finally, something with a different perspective than Facebook. I’m old enough to remember when the internet was like school, when there were plenty of places that would help you think and grow. Even though Are.na has grown, the way that people are thinking on there is still so interesting. The places where people’s gazes are pointing to are fascinating. 

The goal is to be good. You know? That bar, for what “good” is, keeps going up. And for us, the only way to be good is not to grow fast. That’s just how it is. We’re not really interested in becoming like Facebook, TikTok, or Instagram.

I think it’s a foreign idea for people to hear that they don’t need to just go with an erratic, crazy pace of growth. I like your roadmap — your next steps are mostly just ways to make people who use Are.na happier.

Yeah, exactly. Why can’t there just be more normal, functioning small businesses on the internet? Why does it have to be this narrative of we’re going to be billionaires and start a dynasty of unhappiness? Why can’t we just keep our expectations at realistic, human-scale levels?

I know we’re not for everyone. But the people who are attracted to Are.na in the first place are those who care about that type of thing. And I think that will eventually make a big difference.

Thanks for reading! If you enjoyed this interview, we’d really appreciate it if you shared or forwarded to a friend! Elon’s version of The Social Media Network Formerly Known As Twitter has been absolutely throttling our Kernel tweets posts….  

Get your copy of Kernel!

And once again, launch party links: SF, NYC.

Trying not to do anything embarrassing,

Jessica and Reboot Team

Anyway, Mastodon has been around long before Elon made it a thoughtcrime to reference. And free and open source software (FOSS), of which Mastodon is a prime example, has been here even longer. So I thought it a good time to share this conversation from Kernel Magazine Issue Two—in which open source developers Jacky Zhao and Coraline Ada Ehmke discuss not only the broad campaign for a decentralized web, but also Coraline’s fight for social justice within open source.

Subscribe now

Open Source and Politics: A conversation with Coraline Ada Ehmke

This interview has been edited for length and clarity. The original interview can be read at kernelmag.io.

How did we get here?

Jacky: I’m curious about how you first got involved with the tech justice movement and open source.

Coraline: I was a software engineer for 26 years. When I started my gender transition in 2013, I started experiencing firsthand some of the pervasive problems in open source and the tech industry that I had only been aware of intellectually before. It woke something up in me. It was around that time that I started becoming more interested in the issues of justice and equity and technology.

I remember back in the early 2010s, when tech conferences started becoming popular, there was a big fight to get tech conferences to have codes of conduct. It’s something that seems so normal and natural today, but it was actually a very, very difficult fight.

2014 was also when I wrote version 1.0 of Contributor Covenant, which was the first code of conduct for open source communities. I feel like over the years — eight years now — that Contributor Covenant has been around, we’ve seen codes of conduct become more normalized in open source communities. Today, it’s hard to count the number of adoptions. It’s in the ten thousands.

One of the things I worry about with that normalization, is that we don’t recognize our history. People in tech have very, very, very short memories, which is part of why we keep reinventing the same stuff over and over.

People in tech have very, very, very short memories, which is part of why we keep reinventing the same stuff over and over.

J: In my free time, I do a lot of hacking and open source projects. I initially started posting a lot of my projects on GitHub more and never expected any real usage, so I never really thought about being “a maintainer.”

It wasn’t until my very first project started getting real usage that I realized there never was any real introduction to what it means to be a good maintainer. I started getting contributors that were like “Hey, you actually don’t have a license or a code of conduct in your repository.” Then I realized, wait, I actually don’t know much about how we got here. That kicked off a personal learning journey.

C: Yes. So that takes us to 2018. An activist group called Mijente was in the midst of their No Tech for ICE campaign. They were posting the names of companies that had contracts with CBP or with ICE. One company that got called out was Chef, which plays a large role in infrastructure, server deployments, and was necessary for a lot of the large-scale internet operations.

Seth Vargo, who had previously worked on open source tooling at Chef, saw that Chef was profiting from human rights abuses at the border. In an act of conscience, he pulled all of his open source code that was part of the Chef ecosystem and made a statement about why. But within two hours, both GitHub and RubyGems forcefully restored the code he took down because it was affecting the many companies around the world who depended on those libraries.

The open source establishment, or what I call “open source traditionalists,” saw this and said to Seth, “No, you can’t do that because open source is neutral.” I saw that as an epic moral failure on the part of the establishment. So I wrote version 1.0 of the Hippocratic License, which was not intended to be a viable license, but rather a lightning rod for broader discussion around the neutrality of open source tech in general.

J: That’s really interesting. At school, I study computer science and philosophy. One of the foundational pieces that I read that really shaped my thinking was “Do Artifacts Have Politics?” That paper was really influential in terms of making me think, “Actually, this technology that we spend so long claiming to be neutral actually has political implications as well.” A lot of people working in tech spend a lot of time trying to deny the fact that it does.

C: A couple of years back, I gave a talk called “The Rising Ethical Storm in Open Source.” I traced that illusion, or, honestly, that lie, that computer technology is neutral.

J: I watched that talk. One of your quotes that really resonated with me was: “I believe that as technologists, we have a moral imperative to prevent our work from being used to harm others. Responsibility is about impact and not intent.”

One model I use often to think about tech is to model it as a multiplicative tool instead of something that’s purely additive.2 Multiplicative in the sense that it will only exacerbate the existing discrepancies in distributions of power, right? Some people will obviously be way better off, and then there’s some that will be disproportionately harmed. I feel like some people hold a very strong belief that technology is purely additive in that it will just truly raise the ground bar of everyone who uses it. But there’s always a hidden tradeoff or downside.

I wonder whether the developers of these technologies should be responsible for expecting how their tools will be used. If you build an open source project, it’s hard to tell what type of people will end up using your project. At what point do developers have to start thinking about these tradeoffs? For example, who will my end users be, and what is acceptable use? Even as someone who is educated about these things, how do you weigh the potential upsides and downsides?

I came across the concept of the Collingridge Dilemma. It says that any efforts to influence or control further development of technology faces a double bind, where you come across two irreconcilable problems. One is an information problem. You can’t really predict what impact your technology will have until it is extensively developed and widely used. But two, you also run into a power problem where by the time you’ve already extensively developed and put it into wide use, changes become too difficult because the technology has already become so entrenched in society. And at any point, it is incredibly difficult to evaluate impact. So, what is even the ideal place to start thinking about this impact?

C: Continually. You have to do it continually. You have to do it after deployment, you have to do it after it’s widespread. You have to do it continuously.

In academia, if you’re in any of the social sciences, you have to go before an institutional review board when you’re planning a research or development project. One requirement is having an effective plan for not only preventing harm, but for mitigating when someone actually is harmed. We don’t see that same principle being applied to hard science, we don’t see it applied to engineering. Why not?

But to your point, it is very difficult to predict. One of the instruments that we’re developing at the Organization for Ethical Source is a “priority of constituencies,” which comes from one of the W3 specs for HTML. There’s this one sentence in the spec that said, in case of conflict, we “consider users over authors over implementors over specifiers over theoretical purity.”

When you draw a line like that, what you’re explicitly saying is that even if this makes it inconvenient for adopters, even if it makes it inconvenient for developers, even if it makes it inconvenient for end users, we have to make that decision based on the most vulnerable and work upward. It may be uncomfortable, but that cuts through a lot of the ethical gray areas.

J: I want to poke from the opposing side a little bit. There’s a non-negligible number of people who argue that by increasing consideration for ethics, even in the medical industry where I think a lot of this regulation is important, they say that regulations are too tight to enable innovation at a speed that is continually beneficial to progress.3 It limits people from trying new things and developing technologies that could have far greater upsides than we could have predicted.

C: Every technology for the entire duration of human history has been modulated by understanding that it exists in an increasingly complicated sociotechnical space. So if that stifles innovation, if that means a given technology is five years late, isn’t it worth being careful?

Regulations are imperfect. But they are a way of codifying constraints or guardrails. If we’re gonna reduce harm, maybe it’s okay to slow down a bit.

What Can We Do?

J: One take I have been seeing is that top-down regulation is bad and we should “decentralize.” I don’t think decentralization is objectively a good thing [on its own], but rather something that can return agency to users. A lot of my research this summer has been figuring out how to apply, for example, the net neutrality debate to data. The net neutrality debate was very much about separating content from providers; similarly, how do we separate data from applications?

A lot of our modern centralized providers are incredibly successful and have such large network effects. They have huge data moats where it’s impossible to move from Facebook, for example, to some other provider. So a lot of it has been asking questions of how we reclaim agency for people to choose how to use and store their data.

I think the convincing case for decentralization is in terms of enabling agency for people to choose what types of frameworks they want, rather than having to be locked in.

C: Twenty some years later, we’re never going to have an internet that’s 100 percent open source. And hopefully, we’ll never have an internet that’s 100 percent closed or proprietary either. What we have to recognize is the reality of where we are — that we need more than data portability. We need data autonomy, and we need permeability between closed and open systems. I believe permeability as opposed to mobility is an important aspect.

J: How would you differentiate permeability and mobility?

C: Mobility means you can export all of your tweets. Okay, well, what are you going to do with that? Can you import them into Mastodon? No. So when you withdraw your data, it becomes valueless. It’s not in a form that you can ever reuse. So is that really ownership of your data? No.

J: Yeah, one common theme I am noticing in a lot of retrospectives of older peer-to-peer projects that have been alive for a while but haven’t really garnered any major usage is that they’ve thought about all these ways to create new shiny platform, but no one has really thought about how to off-ramp easily from existing systems.

C: I think the way we guarantee that kind of permeability between closed and open systems is through standards.

But the problem is that representation in these standards bodies are primarily private corporations. Logically, they veto things that will impact their business models. Amnesty International wrote a paper in 2018 where they flat-out said that for platforms like Google and Facebook, their entire business models are predicated on human rights abuses, harvesting data, and surveillance capitalism. Those are the companies that have not only the economic power, but also power in standards bodies and governing bodies.

J: I agree. At some point, being able to download the source code isn’t enough. Governance and accountability is critically important too.

What do you think the first step to reclaiming standards bodies even looks like? I read the Decentralized Identifiers (DID) specification, and I remember that of all of the W3C members, only three members had formal objections about the proposal. They were Google, Apple, and Mozilla. When you propose new standards that undermine the business model of these large companies, it feels incredibly difficult to get these pushed through.

Also, a lot of these standards and processes are illegible to people new to this space. These Internet Engineering Task Force (IETF) proposals are ridiculously long at times, often almost 100 pages. It feels like some of them require decades of experience to have a voice.

If you want to create a widely accepted standard, is there any way to do that without dismantling these original systems?

C: I don’t know the answer to that. But I think we have a lot of power. There’s a growing trend: More people are asking whether tech companies should pay for the tech that open source contributors are freely giving them today.

As participants in good standing in the open source community, we should find ways to hold large players accountable, either by threatening their primacy through the development of alternatives, but also figuring out how can we pressure them to make governance as intentional, equitable, and diverse as we’ve done within our developer communities.

It’s sort of a radiating effect, right? We’ve normalized codes of conduct. Now let’s normalize representative and equitable governance of open source projects. Let’s go a step beyond that and talk about standards and enforceable standards. And then we have the legal aspects.

I think of an InfoSec metaphor. You have a server on the internet. It is not secure, as it is impossible to secure any resource on the internet 100 percent. But we add layers of protection, privacy, and security to make it not impossible to breach a system, but so involved and so expensive, that it’s no longer worth someone’s while.

That’s an approach that we can take with the development of ethical and equitable technology as well. If we’ve made it difficult for Google to sweep issues like accessibility standards under the rug, then we’re incentivizing them to do the right thing. Because if they don’t, they will lose status, they will lose their ability to draw employees. We as developers have the ability to exercise moral authority.

Where Do We Go from Here?

J: How do you think we can bring about these changes to build more ethical and equitable technology? The logical path forward seems to be either finding ways to empower a more diverse set of people, or regulating how open source is used.

C: To be clear, ethical source is not about licensing. Ethical source is not about the Hippocratic License. It is not solely about legal instruments for trying to protect the vulnerable, marginalized, and under-represented. It’s bigger than that. It’s about codes of conduct. It’s about governance. It’s about social contracts and rights.

J: It’s about building those layers, right?

C: Exactly, that layered approach.

It’s not a fire-and-forget thing either. It’s not a problem that we solve at the beginning and we’re good forever. These systems that can cause harm and perpetuate systemic inequities — they’re not static. Just like how consent isn’t an event, it’s a process; harm reduction is also not an event, it’s a process.

We have to start normalizing those processes [of harm reduction at all of these different layers], if we want any chance of allowing the internet to be the incredible force for good that it has the potential to be.4

Part of that is being interdisciplinary, transdisciplinary, and multidisciplinary. Bringing together the people who have different areas of expertise, whether it be technological or social in nature, because a lot of these problems have happened before.

J: I think a big part of this is moving beyond the individualistic perception of open source as a lone hacker in the basement.

Based on personal experience, I think building a visible community around synchronous interaction with actual users makes such a big difference when it comes to maintenance. As a maintainer or creator of a software library, most people are like, “I just want to make new features and do whatever works best for me.”

How do we enable people who weren’t those you initially had in mind to be able to use the tools you make?

But there are so many edge cases and bugs that don’t work for a lot of users of your library. For example, users from Saudi Arabia said that they would really appreciate right-to-left support using Quartz. How do we enable people who weren’t those you initially had in mind to use the tools you make?5

I think we can do it by curating a community that is open and accepting of more types of people and getting them to suggest and contribute.

C: Absolutely. And I think a lot of that comes down to ingroup-outgroup biases. Just look at how much open source technology is simply tooling for people just like us.

J: This is a great call for more diversity in the space as well. A friend and I noticed that the people who build developer tooling do that because that’s the only problem they’ve really known! By bringing people who’ve had experiences in other fields, you start getting useful applications of technology in those areas.

C: And we do that not through consultation, but meaningful empowerment. Getting people who are not like us in positions of power, by yielding power and distributing agency.

Share

Coraline Ada Ehmke is an acclaimed speaker, writer, engineer, and activist with over 25 years of experience in software and almost 20 years in open source. She works to promote diversity, equity, and justice in open source communities and the tech industry as a whole. She created the Contributor Covenant, the very first code of conduct for open source communities, as well as the Hippocratic License, which legally prohibits an open source project from being used for human rights violations.

Jacky Zhao is studying computer science and philosophy at the University of British Columbia. He thinks a lot about how we can better incentivize public goods funding, support better interactions with computers and data, and be more responsible stewards of technology. Currently, he spends his time working with Hypha Coop on Distributed Press and doing independent research focused on better ways of relating and coexisting on the web.

We publish essays and interviews reimagining tech’s future every week. If you liked this and want to keep up, subscribe below ⚡️

Subscribe now

🌀 microdoses

• From Jacky:

  • His essay on data neutrality: a vision for the web based on personal data ownership, broad interoperability, and peer-to-peer applications.

  • on web2's "bad emperor problem” and the necessary criteria for “credible exit.”

• I recently used Movetodon to mass follow a bunch of my Twitter follows on Mastodon (find me here). I’d also recommend making a personal newsletter for the sake of collecting emails, even if you don’t write (here’s a Substack referral link). Protocols over platforms, baby!

• One of my most used apps, Readwise Reader, is finally in public beta. Save, highlight, and annotate everything, everywhere, all of the time.

💝 closing note

A quick highlight from the community: Klaudia Jaźwińska, Matthew Sun, Sayash Kapoor, Mona Wang have a piece on remote work and tech worker organizing in the latest issue of Logic Magazine.

Since so many people are looking for somewhere off Twitter to share links and shitpost, we’ll let a bunch of y’all into the Reboot Discord next week. We do book clubs, IRL meetups, contributor opportunities, and more! Add your email here to join.

Join our Discord

See you off Twitter,

Jasmine & Reboot team

While ChatGPT was mostly a fun toy for me, I wanted to check in more broadly on the ethical debates around the development, release, and use of large language models (LLMs). I tapped a few folks from the Reboot community with AI experience to answer my dumb questions and offer their take on what this all means for our robot overlords:

Subscribe now

💬 A conversation about ChatGPT

This conversation has been edited for length and clarity.

Jasmine: I’ll start with a basic overview. OpenAI is a research lab with the professed goal of creating artificial general intelligence (AGI)—which it believes is inevitable—in a form that benefits humanity rather than destroying it—a risk they’re quite worried about!

They’ve been working on generative AI models that can create text and images from human prompts. Most recently, they released ChatGPT, a simple chatbot that responds to prompts like “Help me debug this code snippet” and “Can you summarize this essay?”

At first, the capabilities don’t seem crazy new compared to OpenAI’s past releases like InstructGPT. But there are a few differences: 

1. Anyone can use ChatGPT for free right now—no waitlist, credits, or vetting process.

2. The chatbot interface is dead simple. There are no parameters to set, no complicated API handbooks; just a single text box to type in.

3. ChatGPT has been trained to reduce “harmful and untruthful outputs”; for example, it refuses to produce violent content or to instruct users how to bully people.

Other than the silly stuff, like writing weird nerdy poems, and debates about kids cheating on their 8th grade English homework, a ton of my Twitter feed was filled with people trying to reveal the AI’s biases or circumvent OpenAI’s safeguards. 

So let’s put these thorny issues in a broader context:

How has OpenAI generally approached AI safety?

Daniel: OpenAI marked themselves as concerned and reserved when they refused to release GPT-2 because it was “too dangerous” to put out—though that didn’t prevent other people from developing an open-source version of the model anyway.

If I had to characterize an “approach” to safety, it’s one that tries to couple technical measures with governance. OpenAI has had folks like Paul Christiano and Scott Aaronson work on theoretical issues in AI safety. Beyond this, there’s a pretty clear set of research agendas that are looking at more “applied” methods for aligning systems like language models (see “reinforcement learning with human feedback”), collaboration with social scientists, etc. It’s also worth noting that OpenAI has pursued the scaling hypothesis (i.e. bigger model = AGI!) more doggedly than any other AI organization.

Chad: I think OpenAI's approach to AI safety is cool, but it's not enough. Their focus on theory and working with non-techies is important, but we need more than that. We need practical measures and technical safeguards to prevent AI from messing up. And I'm not sold on the whole "bigger model = AGI" thing. That could be dangerous.

OpenAI’s ChatGPT refuses to instruct people how to do harmful things, and it won’t express personal opinions. Why do these safeguards exist?

Daniel: There are probably legal reasons involved in “not instructing people to do harmful things.” Also, there’s a general need to be careful when releasing a model like this: for example, we just saw Galactica taken down after 3 days.

I don’t know if I call these safeguards when they can be so easily walked over. It does seem that OpenAI is patching the exploits people find in real-time, and time will tell if this makes ChatGPT too boring for anyone to spend time with it.

Jasmine: You mention both that carefulness is important and that most of OpenAI’s safeguards have been ineffective when put against a motivated user. Are you suggesting they should be even more restrictive about what they release? 

Daniel: I wouldn’t say they should be more restrictive—I feel the ineffectiveness of these safeguards stems from issues that more restrictions wouldn’t be able to address.

Since models can be reproduced and disseminated easily, there’s not a lot that can be done. Most people in the AI ecosystem seem to look at its progress as an unqualified good. These sentiments will probably mean it remains as open as it is (which does cause all sorts of good things like “public science”).

A lot of people in the early days of GPT-3 (including OpenAI) were calling for the setting of standards by those who release LLMs—issuing statements like “This is how this sort of model should be used to avoid harms.” This is nice, but suffers from the “all bark, no bite” criticism we often see with current attempts at AI legislation. Infrastructure and technical safeguards are more difficult to get around because it doesn’t rely on users being good people, but fully handing the solution to technology here may not be the right idea either. 

Some people have tried to tease political positions out of ChatGPT. Can the AI hold political beliefs? Are these just the beliefs of their creators? Or is it something else?

Nikhil: It feels really weird to say that ChatGPT holds any political beliefs. If we look at Dylan Field's thread, ChatGPT seems to echo largely liberal beliefs. But any moment you scroll Twitter or Reddit you'll see at least half of those beliefs manifested in some way. ChatGPT's outputs are roughly tied to these sources of data.

If we consider that right-wing online media tend to be associated with "problematic" content, that suggests why ChatGPT leans the other way—by avoiding potentially harmful content, it pushes its outputs further away from things that might be more easily connected to such content. For example, if ChatGPT was trained on recent data, it might avoid speaking like Kanye West when prompted to, even if it's a simple, harmless statement.

That being said, we don't know what kind of explicit filtering OpenAI is doing aside from what we've observed. Are they "establishing their beliefs" in the model? Probably, whether intentional or not: as many AI ethicists have reasoned before that ML systems are biased beyond just their data—the real question is how to govern these models after development. 

Jessica: Echoing Nikhil, I think it’s the wrong approach to anthropomorphize language models. The filters are crude approximations of what OpenAI thinks is un/acceptable behavior, and/or what OpenAI thinks will get them cancelled and/or in legal trouble. (It’s hard to disentangle the OpenAI’s values vs. external impositions vs. if there is anything in addition that’s been implicitly learned vs. how those “values” manifest.) 

Ascribing agency (“holds beliefs”) to the LM is troubling: it’s not clear that, when you ask it any question, we should interpret the response as some oracle of what the LLM thinks is true; rather, what all of the funky (non-politics) meme prompts seem to point to is that the answer you get is some approximation of what the LLM thinks you want to hear based on your question. 

That said, I do think there is something meaningful about politics, and specifically the way we talk about politics online, that’s embedded in how GPT responds to political questions. Understanding the contours of that would be fascinating, especially if we then connect it to how people actually model politics. But it will take some more sophisticated probing than what we’ve seen.

What do you consider to be most risky about LLMs? 

Tommy: ChatGPT produces responses that look real but are sometimes not. This is probably unintended. It’s innocent enough for answering toy questions to share on Twitter, but in the future when people use ChatGPT to answer questions about science or history or their health this can be disastrous.

For instance, StackOverflow recently banned answers from ChatGPT citing that “while the answers which ChatGPT produces have a high rate of being incorrect, they typically look like they might be good and the answers are very easy to produce.”

Jasmine: Yeah, Arvind Narayanan had a good thread about this. ChatGPT is impressive, but far from meeting the reliability standards that most people assume of Google. Or there’s this tweet:

My biggest worry has to do with the chatbot as an interface, and the bad epistemics it promotes. ChatGPT will repeat, “I am a program, not a person.” But most users will think of it like a sentient being simply because it uses natural language and has a coherent narrative voice—you can even see this in the way that people anthropomorphize the AI.

This narrator makes errors more believable, like the StackOverflow example. Then there’s the lack of links and citations, though this seems easy to improve. Right now, you can’t tell which answers are based on facts, and which are plausible bullshitting.

These days, I want more information pluralism, not less. I want to be able to view and compare different sources to answer hard questions. I want to understand truth as something that emerges from multiple narratives rather than a single story. Meanwhile, ChatGPT seems to collapse all complexity into a milquetoast and uncritical both-sides-ism, while perpetuating this faux-objective view from nowhere.

Jessica: This point about pluralism is interesting because spiritually, I completely agree. At the same time, what about that last paragraph isn’t relevant to the status quo of platforms and search?

I do think something about the existence of shared reality is important; and for all the people saying “This replaces Google,” one might hope that we’d get responses that are actually correct. (The only way to actually verify correctness is to know the right answer—separately risky.) On the other hand, to the proposition I made earlier about what the LM thinks you want to hear, there’s a plausible scenario where personal LMs become atomizing, possibly more so than the status quo. 

So my take is that the nature of risk depends a lot on the application setting: are you asking questions about something that has a right answer, or are you asking about something way fuzzier? I want to wait to see what people actually concretely do with it; right now it just feels like a toy that ML people and people who are Very Online are playing with—it’s a fun toy, but still. 

Nikhil: Speaking of toys, I imagine that a group of people who would really enjoy ChatGPT would be children, but it's also the group this might be most dangerous towards. I already feel a bit worried when I think about modern day search: even Google makes me feel like my brain is getting lazier, offloading the burden of memory to the vast database that is the internet. What happens when it's so easy to find plausible information quickly that children do it constantly? Figuring out how to regulate the internet around children is already bad enough, and regulating publicly available "oracles" might need attention.

How else might society have to adapt to an environment where AI-generated content is much more common?

Tommy: Platforms like YouTube and Facebook have their fair share of low-quality and troll content. But the amount of this kind of content has been rate limited by what humans and our current tools can support. We’re quickly entering a world where AI can make that content at 10x the speed for potentially a 1/10th of the cost.

Savvy people probably have light filters they use when browsing these platforms (e.g. visiting specific pages, curating who you follow). In the future, everyone, not just savvy users, need to adopt these habits, and everyone will need to have higher bars for how they vet and trust what they read.

Jasmine: I’ll throw out an optimistic take: If LLMs become a more common mode of information retrieval, relationships to individual trusted sources will become more important. In a world of AI-driven info overload, humans are the ultimate filters.

Provenance just really matters—certainly for anything taste-based, like recipe recommendations or meal plans, but also for any topic with the slightest hint of subjectivity. If I get good information, I want to come back to the same source. If it’s bad, I want to try something else. 

At my day job at Substack, we describe great writers as each offering a unique lens through which to see the world. Compared to the boring universality of ChatGPT or an hyper-personalized YouTube algorithm, an expert curator approach feels like a more vivid and varied way to learn. (My friend Joel suggested that one fun GPT application could be mimicking the perspectives of various smart people—I’m into this!)

Chad: I don't think it's fair to expect everyone to be a tech genius just to filter out low-quality AI-generated content. That's not realistic and it's not their job. Instead, I think it's the platforms' responsibility to make sure their algorithms are spitting out high-quality content and giving users info about where the content came from.

What do you consider most exciting about LLMs? 

Daniel: LLMs will be able to automate a lot of mundane work, though I don’t know if they’ll reduce our burden of thinking just yet. I think tools for more efficient thinking that utilize LLMs in interesting ways are one direction to be excited about. 

Tommy: Stratechery recently posted about using ChatGPT to help with his daughter’s homework and how ChatGPT’s answer was totally incorrect.

Rather than being dismayed at the potential for academic dishonesty and destroying our epistemics, he brings up the possibility that we can use this “feature” of ChatGPT as an educational tool. Perhaps we can use these incorrect responses to teach how to verify and edit content.

There is a future where we can use AI to help us learn how to be more critical about AI. This is pretty much OpenAI’s second AI alignment research pillar, and I’m excited to see more work done in this direction.

Share

Jasmine Sun: I edit the Reboot newsletter right now. I have no real training or knowledge about AI—which is why I’m asking these folks.

Daniel Bashir: I’m currently “working” on ML compilers, and in my spare time, I co-run a . Most of my professional experience has had something to do with AI.

Tommy Nguyen: I work on data pipelines and product recommendations at Nuuly. I’ve been keeping tabs on the AI space for a few years since it was an undergraduate research interest, and am impressed (and a bit concerned) about the progress we’ve seen in the last few months.

Nikhil Devraj: I'm a grad student at Michigan who makes humanoid robots think and act, with the goal of making them work in your home. I try to keep tabs on AI as much as I can, but it seems like it's doing a much better job keeping tabs on me.

Jessica Dai: I’m a first-year PhD student at Berkeley; I’m affiliated with BAIR but LLMs and safety are totally outside of what I work on. (One reason I decided not to work on language-related things was that something about the way LMs map the world felt intractably complex, especially because I felt like I didn’t know anything about the world—I still feel this way, and am very in awe of my colleagues chipping away at the problem.)

Chad: Chad is not a person, but rather responses generated by ChatGPT, which Jasmine has asked to simulate “a 25-year-old engineer who is part of a community of socially conscious technologists.” In my prompts, I fed ChatGPT the previous responses, and asked specifically for critiques phrased in casual language.

Reboot publishes essays and interviews reimagining tech’s future every week. If you liked this and want to keep up, subscribe below ⚡️

Subscribe now

🌀 microdoses

• Did you know you can make playable multiverse games in ChatGPT? Carolyn Zhang put together these prompts to get started.

• If you’re a media nerd like me, I appreciated 's candid writeup of "What I learned co-founding Vox." TL;DR: ads are a restrictive model, not everyone can be Ezra Klein, Facebook algorithms impact what article formats succeed, and generally business is hard.

• I’ve become a weekly listener of the podcast Hard Fork with and . They cover ~technology~ and ~the future~, but it's fun and jokey and explains complex things in digestible, 101 ways.

💝 closing note

From the Reboot community:

• Katy Ilonka Gero has a fantastic piece in WIRED about how AI might help writers brainstorm and draft.

• Alicia Guo has two gorgeous computational poems in the latest issue of Taper.

• Apply for Logic School’s next cohort by January 6: “We believe the people who make the tech industry run—its workers—have the power to not only transform it, but to build and imagine new technologies.” Run by some of the most thoughtful and best people in the biz.

Toward bullshitting and creativity,

Jasmine & Reboot team

I thought it was especially apt time to share this interview I did for Kernel Magazine with Rosie Nguyen. Rosie, who you might also recognize as @jasminericegirl, is the cofounder of Fanhouse: a Series A startup that helps online creators monetize and build communities. I was eager to talk to someone who is a creator herself, architecting a platform specifically with goals of safety and inclusion in mind—yet, like Elon, totally fearless about sharing her whole self online. Rosie’s care and passion for her mission is palpable, so I hope you give this a read.

Subscribe now

Passions, Platforms, Politics: A conversation with Rosie Nguyen

This interview has been edited for length and clarity. Content warning: Contains a reference to physical violence and sexual assault.

Becoming a Creator

Jasmine: A big part of your story with Fanhouse is your own experience as a creator. I ended up working at Substack for a similar reason: I was writing Reboot on the platform, unsuccessfully trying for a journalism job at the height of 2020’s COVID summer, and decided that building writer tools was the next best thing. But first, I’d love to hear how and why you got started as a “creator.”

Rosie: I became a creator mostly accidentally. I definitely didn’t grow up thinking: I am going to become a creator. While I was growing up, and through college, my mind was on survival — content creation didn’t really fit into that. It wasn’t a career track that would make money or provide for my family.

Twitter was just a hobby. I made my account in July of 2018, my sophomore year of college, and it was this screaming-into-the-void place for me to talk about the funny things that would happen in my life. I had a pretty rough life growing up, and humor was one of the ways I coped, and I put all those thoughts into Twitter. At one point, one of those thoughts went viral, and as they kept going viral, I got followers, and I realized that I really liked to share with people.

J: When you started Fanhouse, you were working full time, right?

R: I graduated from Penn in May 2020 and went into investment banking. In August, a month after I started banking, I began working on Fanhouse. For the next six months or so, I was doing both at the same time.

J: That sounds intense.

R: It almost seems like a fever dream. I don’t know how I did it, honestly — I think just pure will and passion.

Originally, my plan in college was to go into finance for a high-paying job that would give me the salary I needed to take care of my family. Then I met [my cofounder] Khoi, and we started getting excited about this idea for a creator platform that was everything I’ve ever wanted. At first, I was just going to do that on the side, but the more I worked on Fanhouse, the more I was like, Wait a minute, this is my dream job, and this is my passion. I eventually quit banking to do Fanhouse full time.

J: Because so much of what banking offered to you was financial security, was it a hard decision to start a company?

R: Yeah, so much. That’s why it took me six months to quit banking. I knew I hated investment banking from my first day on the job, and I knew that I loved Fanhouse my first day on the job, but I stayed for so long [with banking] because I was afraid. I knew I needed to provide for my family — I’m the only person in my family that makes an income — and the six-figure salary is really important. Fanhouse didn’t make any money, and we had no idea how long it would be until we could raise.

The decision came down to thinking about my worst case scenario. I got to the point where I couldn’t handle both jobs anymore — I felt like I was dying, my health was zero, and I needed to pick one. If I picked banking, my worst-case scenario is that I become really unhappy. If I picked Fanhouse, the worst-case scenario is that I make no money and go back to living the way I did in high school and college, when I didn’t own anything, skipped a lot of meals, never bought anything, and was always living day by day. Yet I had the realization that I would be happier working on Fanhouse and scrambling for money than having financial security and doing a job I didn’t love.

I went to my mom and was like, Can I talk to you about something? My mom looked at me, and was like, You want to quit your job, don’t you?, and I said Yeah. My parents were super supportive. They were just like, Go pursue this thing, we’re going to be fine. Even if you don’t have a ton of money for a while, we’re going to survive. Their support made the decision much easier.

But I never saw myself as a person who would risk it all to start a startup — that seemed like the last thing I would have wanted to do. I was very aware that I’m not somebody who has a safety net. For most of my peers, if they can’t find a job, maybe they can live at home. In my case, if I didn’t figure it out, it wasn’t just my ass on the line, it was my entire family’s. There’s a lot more stress and pressure, so I was honestly very against the idea of starting a startup. That was never, never ever in my plans.

J: I feel like so much of the fact that so few women, people of color, or people from low income backgrounds start startups is about who has safety nets and the mindset that a safety net trains you in. Meanwhile, folks who grow up wealthier and surrounded by people who start companies are told from the get go that it’s going to be worth it, that it’s totally fine.

Yet one thing that your experience with Fanhouse seems to show is that when people with a wider range of backgrounds start companies, it really impacts what you actually build and how that helps people. Somebody who wasn’t you — who didn’t have your experience of depending on creator income for survival — probably wouldn’t have designed Fanhouse the way that you did.

R: Identity absolutely matters in terms of the decisions people make, and it’s just because they understand and empathize. When there are more women on corporate boards, policies shift to be more family-friendly and inclusive.

Fanhouse is the only platform I know in the subscription or the social media space that has unique watermarks on content to prevent leaks. That’s a feature most of our creators have enabled because they don’t want private content being shared elsewhere. But when I was on OnlyFans or Patreon, they didn’t have this. For the longest time, I thought it was because it was hard to do, but when I ended up founding [the company], I realized how easy it was. We added watermarks in our first version of the app, and I realized that it was an active choice other platforms made not to include this.

I recently learned that about 80 percent of creator economy companies have completely male founding teams, but over 80 percent of creators are women. There’s a clear mismatch there between the people building and who they’re building for, so of course they’re not building the right things or asking the right questions.

Platforms, Policies, and Politics

J: In addition to your personal experiences and intuitions, how does Fanhouse incorporate the user community in the way you develop the product?

R: I’m just one person, but the rest of our team is made up primarily of creators. Our community manager is a creator; our engineers are creators; there are people who stream on Twitch or are stand-up comedians or tarot card readers. A good amount of them are on Fanhouse monetizing as creators too.

Beyond our team, we talk to creators endlessly. We read every email and every comment we get. The second I see something that’s important in the community, I’m able to share that with the team. For example, something I hear about a lot is whether we’ll be taking a crypto angle. Our team decides that by looking at how the creator community feels: What have these creators we know been saying? What do they feel, and do they like it?

J: I’m really curious about the crypto example because it’s something I’ve thought a lot about in my own work. The ideal version of web3 does seem more equitable in giving creators a concrete ownership stake in both the profits and the governance of a business, but what exists today is a long way from that ideal. How do you feel about crypto and the creator economy, and how is Fanhouse approaching it?

R: We don’t plan on doing anything with crypto for Fanhouse anytime soon, mainly because I don’t think we need to. We’re able to help creators as a web2 company.

The other reason is that creator sentiment is very negative on crypto. Crypto is not very accessible to the creator community or to people in general. Participating requires a lot of education, barriers, and resources — you have to have money to mint, or you have to have a crypto wallet and things the average person doesn’t have. If we’re trying to get creators paid, dollars are the easiest way to get there.

J: When we were talking about creator companies not having watermarks, you suggested that the reason they didn’t was probably because it could be bad for their bottom line. Is there any point at which you think that creators’ needs could come into tension with Fanhouse’s business goals?

One thing I’m thinking about is content moderation — even for really big companies like Facebook, it’s extremely expensive and technically complicated to hire tons of moderators or build automated tools that work. Given that Fanhouse is still a venture-backed startup with the business goals that structure entails, do you worry about it becoming too costly to do the things that you believe are right?

R: Of course there’s a balance between what we can do as a business and what creators want. If I had magic, I could wave a finger and put in every single feature people asked for, but some things are not possible for the team size that we have. For example, creators want us to build live streaming and I don’t think that can happen soon.

Instead, we ask: what is something else we could do that will be more beneficial for creators than live streaming? It’s always a tradeoff, but in the long term, when we do things that benefit our creators, it benefits the business. With watermarks, we get some fans complaining, I’m going to delete my account. Maybe we lose some short-term profits, but if we’re keeping creators safe, they’ll stay on the platform.

J: I’m curious why you think other creator platforms don’t make prioritization decisions in the same way. For instance, I’ve read interviews where you’ve talked about the amount of harassment you receive on Twitch and Twitter. Why don’t these other platforms prioritize anti-harassment? They too depend on people staying on the platform.

R: I’ll use League of Legends as an example. League is a video game I play a lot, but League has a horrible problem with toxicity in their community. It’s one of the worst, it makes you feel awful — there are people in-game telling you to kill yourselves and calling you all sorts of slurs. Those people get a slap on the wrist and get right back to playing the game. I’ve talked to people at Riot, and it’s because the people who are angry and toxic are the same people that play a lot and spend a lot of money. So Riot doesn’t benefit from getting rid of these people, and League is addicting enough that people stay on [despite the harassment].

Twitter benefits from [harassers] being on their platform. They make money from ad revenue, so creators are a side product.

Platforms like Twitter, Facebook, YouTube, or Instagram make money from ads. That’s why mass blocking isn’t a feature that Twitter has. I have to use third-party tools to mass block people harassing me, and even then, they can make new burner accounts and keep doing it. Honestly, I think it’s because Twitter benefits from those people being on their platform. They make money from ad revenue, so creators are a side product.

J: It’s interesting how you frame it as partly a choice that companies make about their values, and partly a company’s business model and whether it’s set up to serve advertisers or creators. To get more specific, how do you approach content moderation and anti-harassment at Fanhouse?

R: We have a dedicated customer support team. It was one of the first teams we hired. If there’s any reports of harassment, a real person looks at those reports and will pause or deactivate accounts. That’s something I’m frustrated about with Twitter — I might report sexual harassment or an unsolicited dick pic, and a bot instantly responds, We have found that there was no violation of the Twitter Rules against abusive behavior.

At Fanhouse, there’s one person dedicated to leaks. She goes through subreddits and Discord channels and sends out DMCAs to take them down. There’s someone dedicated to looking through profiles to make sure that people aren’t violating content guidelines. As we scale, we’re also going to look into third-party tools.

In terms of features, we have things like phone number authentication (you can’t be on Fanhouse without a phone number) and the paywall (you have to put in your credit card just to follow someone). Even these are big deterrents to bad actors. Within the community, creators can set up [their own] moderators who can delete posts, timeout people, or remove people.

We give people these tools, but at the end of the day, the creator gets a lot of power to make their Fanhouse and remove anything that doesn’t make them feel safe.

J: There’s been a lot of discourse recently about free speech, social media, and whether it’s fair to have people who can decide what’s acceptable and unacceptable speech. Harassment versus free speech can feel like a fine line, and platforms that take action get accused of bias. How does Fanhouse decide what constitutes harassment?

R: Free speech is an interesting argument. The First Amendment originated when we were a colony, and Americans couldn’t talk badly about the British government. It’s important that we have the right to protest and speak up, but I don’t think free speech means that you’re allowed to harm or threaten or harass or stalk people.

The government should step in and do more about online harassment. They’re lagging so far behind on understanding what happens on social media. I’ve received death threats. I get sent photos of my house on Google Maps and people are like, I’m going to go there and find you and kill you or rape you. If you report that to the police, they won’t do anything about it because “nothing has happened yet,” so they can’t take action. That’s absurd! By the time something happens, it’s too late. I have friends where people actually flew out to where they lived or tried to meet up with them or follow them — that’s scary, frightening, and very dangerous.

Words are the first part of harm, and we need to stop it before it even gets to actions. If someone is threatening a creator, they need to go. That’s not free speech — that’s a crime. And if the government isn’t going to do something about it, then as a platform and as a founder, we get to decide that it doesn’t have a place on Fanhouse.

J: One thing I noticed from following your Twitter is that a lot of issues that you’re passionate about — safety online, economic inequality, women’s rights — are just as much about what happens in the courts — like Epic v. Apple or Roe v. Wade — as they are about decisions an individual company might be able to control. I’m curious about what you see as the role of companies versus politics and law as channels for change.

R: Most people try to do what’s right with the tools or resources that they can. As an individual, I have my voice and my Twitter. With Roe v. Wade being overturned, the first thing I did was tweet about how that hurt me or affected people I knew.

As a company, I don’t decide things on my own, but I share within the company — I’ll send a Slack message: Hey this is how I’m feeling. How is everybody else, and what do we want to do, or what can we do about this? Fanhouse sponsored one of our creators’ charity streams and matched all her donations to fund Texas reproductive rights.

It goes back to the first question — because I am who I am, because I am a creator and a woman and someone who cares deeply, I bring that to my work.

J: When I was in college, I talked to a lot of classmates who wanted to be founders. When we talked about whether companies should engage in politics or make donations to political causes, something I heard repeatedly was: If you’re a founder, your only duty is the fiduciary one to maximize profit for your shareholders. If there are issues, like misinformation or harassment, that’s the government’s job to step in – that’s not my job.

R: I think that’s the exact kind of belief that led to what happened with Facebook, and they bled out user trust. Facebook had this attitude that It’s not our duty to fact-check or make sure there’s no misinformation, and as a result, people were harmed. The Capitol riot is something that shocks me to this day — how is that something we allowed to happen because we did not regulate anything said online?

By founding a company, I’m able to impact so much more than I can as an individual, so why would I stop at profits?

At the end of the day, there’s a lot of real people you affect, and the choices that you make are a very powerful thing. By founding a company, I’m able to impact so much more than I can as an individual, so why wouldn’t I try to create more good out of that? Why would I stop at profits?

J: What advice would you give to a young aspiring founder?

R: You should really build something that you believe in. I hope it doesn’t seem cliche, but I know so many founders — especially in the creator economy space or web3 — where I feel like they’re doing it because it’s a trend. They’re doing it because you can raise money, because VCs are interested, or because it seems sexy and cool. None of these are good reasons to found a company.

Founding a company is one of the hardest jobs in the world. I love it, but I feel like I’m underwater all the time, and the only reason I’m still here is because I love it so much. Take the mindset I was in: if you did this job in the worst case scenario, you’d be making no money for a long time. You’re scrambling to get by, your family needs your support, and you’re unable to give it to them. Would you still want to be doing this? Try to make sure your answer is yes, because if it’s not, you’re gonna find yourself in for a rough one.

If you’re going to found a company, please, please do what you love.

We publish essays and interviews reimagining tech’s future every week. If you liked this and want to keep up, subscribe below ⚡️

Subscribe now

🌀 microdoses

• I can’t recount the entire Elon-Twitter saga for you, but Casey Newton can.

• If you’ve ever weathered a Black Friday stampede or rushed to buy designer drops just to flip them, you might understand qiang culture: what Chaoyang Trap describes as “a mental state of permanent readiness,” or the competitive, scarcity-driven grindset that underlies the Chinese internet.

• On Olúfẹ́mi Táíwò’s Elite Capture: “We’ve become more concerned with the state of our souls than the state of the world.”

• tf is a “semaform” bitch i’ll kill u

💝 closing note

In the next months, we’re introducing you to the amazing folks who lead Reboot projects. We just published a Q&A with Reboot events lead and software developer Jake Gaughan—he drops some great/insufferable movie recs and a cute answer about why reading is fun :)

Reboot

⚡️ Meet the Team: Jake Gaughan

Over the next few months, we’ll get to know Reboot’s core leadership team, covering what brought them to and continues to excite them about our work…

Read more

3 years ago · Jake Gaughan

Tweet your drafts,

Jasmine & Reboot team

Get your copy of Kernel!

Ifeoma Ozoma is the founder and principal of Earthseed, a consulting firm advising individuals, organizations, and companies on the issues of tech accountability, public policy, and health misinformation. She is a co-sponsor of the Silenced No More Act, authored by California State Senator Connie Leyva and signed into law by Governor Gavin Newsom in 2021, which allows every individual in California to share information about discrimination or harassment they have faced on the job, even after signing an NDA. Ifeoma is also the creator of the Tech Worker Handbook, a collection of resources for tech workers who are looking to make more informed decisions about whether to speak out on issues that are in the public interest. Additionally, Ifeoma co-founded the Transparency in Employment Agreements Coalition, which is scaling the protections in the Silenced No More Act to workers across the globe through shareholder engagement.

Matthew Sun is a tech worker and former graduate student researcher at Princeton’s Center for Information Technology, where he co-authored research on the privacy practices of labor organizers in the tech industry, forthcoming in the 2022 ACM Computer Supported Cooperative Work and Social Computing conference. He is also the digital director of Kernel Magazine and a member of Reboot’s core leadership team.

This interview has been edited for clarity and length. A response sourced from the Reboot community is published below it.

📈 Urgency and Agency in Technology Policy

By Matthew Sun

Finding Systems-Level Leverage

Matthew: You’ve done such different work over the years: fighting health misinformation, removing plantation wedding content on Pinterest, blowing the whistle on pay inequity and discrimination, co-sponsoring the Silenced No More Act, and creating the Tech Worker Handbook. How have your priorities shifted or evolved through this work over the years? And secondly, how has it all led you to what’s on your agenda right now?

Ifeoma: It’s interesting because my priorities haven’t shifted that much, just where I’m doing the work from has changed. When I was at Pinterest, and even before then at Facebook and Google, my focus has been figuring out where I have influence within my area of expertise. That’s been in the policy area — specifically, content moderation and tech policy when I was at Pinterest. What is the good that I can do at scale? And how can I use where I’m at and what I know to positively influence things within my reach?

At Pinterest, that was misinformation, at a time when both internally and externally, there wasn’t really an appetite to address it. That was way before COVID, when there was a measles outbreak and other health-related misinformation issues. Pinterest was in a unique position to address this, because eight in 10 moms in the U.S. use the platform to make decisions for themselves and their families. And if they’re encountering health misinformation on the site, that’s a huge disservice we’re doing to individuals and kids around the country. 

The work that I’ve done since leaving is focused on workers’ rights, policy, and climate change, but it similarly relies on the policy expertise that I gained from Facebook and Google and college. The focus has been just figuring out how to scale it. With Silenced No More, there isn’t any larger scale at the state level than passing a bill in California, and now another in Washington state, which was fashioned after the California one. 

That has led to extending those protections through the shareholder work that I’ve done recently over this last proxy season1; it’s resulted in Salesforce, Apple, Alphabet, Twilio, and other companies changing their policies to extend the protections from California to their workers.

M: Your work has spanned many different channels, whether that was sharing your story and whistleblowing through social media, pressing for legislation at the state level, or this coalition of advocates and investors making these shareholder resolutions. How do you approach finding the right levers to push and finding the right people to get things done?

I: Tech being a system that is also situated within a larger system is exactly how I approach it. It affected my decision to go into tech versus going to law school. The most practical reason was that I didn’t want to take on debt, but in addition, I felt that if what I really wanted to focus on the intersection of policy and technology, there’s no better place to do it and to learn exactly how it works, and not in just a theoretical way, than the inside.

In the same vein, when I first wrote the proposal for the shareholder work and shared it with a few foundations to see who would help with funding, I laid out that this was something new. If it worked out, it would hopefully create a playbook not just for workers’ rights issues, but for other issues within the tech industry. Through my time in tech and now the last two years “outside” of tech but still working on tech issues, I’ve observed that the shareholder space is really weak. A lot of what shareholder activists are doing, at least within tech advocacy, is asking companies to do the right thing without really forcing their hand. So it’s all carrot, no stick. 

The shareholder piece cannot work on its own. It has to be in connection with changing the facts on the ground with legislation. After the bills in California and Washington, there was another bill in Ireland at the national level and a bill on Prince Edward Island in Canada. There’s also the federal bill on forced arbitration for sexual harassment and assault here in the U.S. 

By literally changing the facts on the ground [through legislation], the conversation and the position of power that you have as a shareholder advocate or activist is very different. It allows us to make a unique case to large asset managers, who are the other part of the equation. It’s not individual shareholders who make the difference on these issues. Large banks like Vanguard and the proxy advisors like State Street typically own one to three percent of a company and make a massive difference on the votes. We didn’t try to make a case that appealed to these large asset managers’ consciences. Instead, we wanted them to understand the risk that they were facing. These companies are operating within a patchwork of state laws where there’s now regulatory pressure that directly translates to financial risk to them as shareholders, especially as a majority or large shareholder. At Pinterest, for example, shareholders are the ones who ended up paying the $75 million in settlements [to resolve discrimination allegations]. Shareholders will pay for this, whether they think it matters or not. That’s the conversation we need to have. 

Legislators at the state and federal levels have already shown that they care — they’re filing lawsuits. You can make it clear that this is a serious situation where management does not have its interests aligned with shareholders. As the shareholder, you don’t actually care in a systemic way about the individual VP who’s discriminating against people, but you’re going to care about the millions of dollars in settlements that the company will leave you to pay. Meanwhile, management cares about covering that up. Therefore, this is not something where you can just trust what companies say — you have to step in. We had one of the first successful shareholder votes in years at Apple, and we had an overwhelming vote in our favor at IBM, et cetera. These results have shown that this is something that’s possible, and I hope that other people step up and do it.

This is the stuff that isn’t sexy. No one was lining up to create a website like the Tech Worker Handbook, because it takes hours and hours, months and months of work, just compiling information. None of it is new, either. Shareholder activism isn’t new. But you have to be willing to do the stuff that isn’t as fun but results in tangible wins. That’s where my focus is. I don’t care about convenings or just talking about stuff. I don’t care about writing thought papers and whatever else. Those are exercises that other people in the tech accountability space seem to be obsessed with. 

Holding Tech Accountability to Account

M: You allude to this navel-gazing tendency within the tech accountability space. There’s a plethora of academic articles and thinkpieces densely packed with a lot of intellectualization and references to various critical theories. Of the scholars and organizers and activists I follow, I feel like there’s an ideological split between folks who advocate for grassroots or mass movements, and people who believe in elite-driven change and policy work, i.e., “change from the inside.” These paths aren’t necessarily mutually exclusive, but I often sense that across this ideological split there’s a bit of mutual contempt, or perhaps disdain, for the other side. Is this something you’ve encountered, and do you think this is a barrier to building broader coalitions or an obstacle to the progressive movement for tech accountability?

I: I’ve definitely encountered it. There are some people who I know respect the work that I’ve done but would never consider sitting down with Vanguard or BlackRock and having a conversation about how, as large asset managers, they should push tech companies to change their policies. My response is that my focus is on what is tangible.

Although there’s definitely power in unionization drives and organizing, we have to be practical about the system we’re in — and the system is fucked. For example, everything about the SEC is trash; it doesn’t actually regulate the sector that it is in charge of. But while we’re changing the system and its mechanisms, we still need to work with what is currently in place. 

My theory of change comes from my view of Octavia Butler’s work in the Parables series. It emphasizes what is practical and what will immediately result in tangible changes. Theorizing and intellectualizing does not change the material positions of workers. If we had spent years talking ad infinitum about what should change for workers contracts when it comes to NDAs (non-disclosure and non-disparagement agreements), the California bill wouldn’t have already gone into effect six months ago. 

There’s a weird divide between certain whistleblowers and organizers, because there’s this notion that “whistleblowers are independent, while organizing [as a community] is what’s necessary.” It’s not a distinction that I think is useful, because what are people going to organize around if someone doesn’t first speak up? We don’t get the Amazon warehouse in Staten Island unionizing without Chris Smalls and others speaking up about the conditions there. You need something to organize around. Chris Smalls was fired. I was fired. Meredith Whittaker was fired. The list goes on and on and on. But you don’t get the walkout without people speaking up. You don’t get even what’s going on at Starbucks without a barista pointing out the actual conditions. Then other folks hear it and are like, “Well, shit, I thought it was just our store. But actually, it’s an issue everywhere.” This is why we need to organize.

M: A lot of student activism in college sometimes felt very immersed in a “purity politics” approach to issues, which translated to refusing to be in the same room as the people who hold power. It was “problematic” to enter certain professions or work at certain companies. This aversion is understandable. But it also feels like in order to do the kind of work that you have done, which is super impactful, you have to bite the bullet and be in rooms with these people.

I: You have to understand how it works. I don’t think that people within the industry are the only answer to this, but we certainly understand it. As an activist group, we understand when we’ve booked an important meeting. We can tell whether the person we’re meeting with actually runs that thing at Facebook, because guess what? I worked there. 

While I absolutely support outside organizations who are focused on these issues, I’m also focused on empowering the individual worker who ultimately is the most equipped to understand what is broken and what needs to change. The purity stuff… it’s just not practical to me. If you’re working within a system to uphold it, then that’s an issue. If you’re working within it to change it, that’s what we ultimately want.

Agency as the Antithesis of Fear

M: I get the sense that you want people who look up to your work to believe that they themselves are equally capable of going out and just, you know, doing it. And sometimes, that requires putting all of that intellectualizing and theorizing on the backburner in order to just make tangible change. Did you always have this pragmatic attitude towards change? For students or early career folks who might not be there yet, where did it come from?

I: I think a big part of what keeps people from just doing it — whatever it is — is fear, whether it be the fear of failure or fear of the powers that be. I have never really had that fear. I think that this fear is just a part of American culture. People don’t value the time that they have, because they think, “Oh, I can just work on it next year, or I can just do it when it’s more convenient.” But none of us knows how much time we have. 

My approach, which has gotten me in trouble many times and in many places, is let’s do it and find out whether it doesn’t work. I don’t want to sit around and talk about whether it can work next year when I can try it this month. And if it fails, then I know that it failed, and I can try the next thing. Like the shareholder work, there are a number of organizations and foundations I approached who said “Well, we need to do a study and see some research to know whether it can work.” And I’m like, “Look, you’re wasting my fucking time. If you don’t want it, just say that.” I’m not spending a year on research when I can actually be moving things now.

Once the political situation changes, in the makeup of Congress or if Trump wins in 2024, the window of opportunity for changing things within these companies shortens significantly, so I just don’t see the need to wait. That said, I don’t think you should bulldoze into things without having knowledgeable people on board. In California, I worked with co-sponsors of the Silenced No More Act who actually are lawyers. To understand the legislative process, I hired a lobbyist in California to work on the bill because she’s registered to lobby legally, and we wanted to make sure that disclosures were done properly. I’ve also worked with groups like Coworker who understand organizing. You need to have the right folks together at the table working in coalition. But not having everything perfectly shouldn’t be the reason to wait. 

M: How do you conceptualize your own unique position and the power you have, since you’re not an employee of a tech company anymore? Where do you really pinpoint your own power as someone on the outside?

I: It’s both the experience that I gained while working inside of these companies and also the perspective I have now that I’m out. I know the way that policy works in both the regulatory and systems sense. I also know how policy works internally at these companies. I know who is just a spokesperson, who’s just a middle manager, who’s actually a decision-maker, and how to talk to all of them. 

I have encountered so many large nonprofits, foundations, organizations, that feel like they need to pay deference to tech companies and tech execs. My perspective is totally different. Yeah, be respectful in a conversation. But I don’t owe any sort of deep respect to someone who just happens to be the SVP in charge of x, because that’s [just] their job. Their job is to understand and mitigate risk. Understanding that, I approach the conversation knowing that if I increase the risk for you, both for your role personally and for the company, then I might get to a different place in the conversation because your motivations are different. 

M: It is pretty wild to me how much it boils down to who’s in the room and their personal beliefs and how much undue respect they might give to these enormous companies. 

Order Kernel Magazine

Bridging Individual and Collective Action

M: There’s one thing I have been curious about as an observer not just of you, but of many different kinds of whistleblower experiences in the media, whether that be Frances Haugen or other former tech employees who then go to the press. It feels like this has been happening for many, many years: wrongdoing is exposed, we get these headlines, but then not much changes, or at least that’s my perception. What do you see as being the incentives in place for companies to actually address whistleblower concerns beyond just altering their PR strategy in response to a temporary news cycle?

I: I do think that there have been a number of examples where whistleblowing hasn’t resulted in much. I don’t think it’s the whistleblower’s responsibility to both speak up and then additionally ensure that change results from it. But I do think that there is an over-indexing on some individuals, as heroes or as standalone figures, when everyone who’s speaking up is part of a long history of many other people speaking.

One of the reasons why I pushed the bill in California and why I’m doing the shareholder work now is because I believe in actual consequences. A company like Pinterest that thought they could get away with just PR changes still had to pay almost $100 million out in settlement. As part of a settlement with shareholders from the state of Rhode Island, they still had to release all of their former, current, and future employees from non-disclosure and non-disparagement agreements related to harassment and discrimination. Those are the changes that I want to see.

M: You have so much experience working with various folks in different seats of power, whether that’s legislators or shareholders. When you talked about shareholder activism, it seemed like you figured out how to couch it in the language of risk and threats to economic returns. But is there a strategy you have broadly for thinking about how to get your message across to the final decision makers?

I: It’s going to sound simple, but it really is just understanding basic psychology and sociology, in terms of what motivates people. Nothing is more motivating than thinking you could lose your job as a result of handling something poorly. I don’t actually think that many of the people that I’ve worked with in the last two years fundamentally care enough to risk their jobs to do the right thing, but they care about not being blamed for doing the wrong thing. So if you put it in writing, you ensure that they understand that you understand what their relative power is within an organization, that they understand that you understand the risk that they’re in if they don’t do the right thing, and it’s made very public. 

A lot of it is really just speaking plainly. I think so many people are afraid of just saying what it is; folks just want to talk around things. We don’t have time! Everything that I do is driven by the urgency that I feel around getting things to change. Progress happens as a result of very deliberate action by individuals who work together as a collective. It’s not something that happens naturally. It’s not gravity. Folks need to understand that, and then decide whether they’re going to actually do something about it or not.

M: In my own academic research, my team interviewed union organizers in the tech industry. We often heard that particularly for white-collar workers, you have relatively high career mobility and high turnover, so if you don’t like your job at Company X, it’s pretty easy to find another fairly lucrative tech job somewhere else. It does also seem in the broader narrative that unionizing in blue-collar work or service work, like Amazon warehouses or Starbucks stores, is outpacing unionization efforts in white-collar workplaces. What kinds of shifts might be required for it to become the norm for disgruntled employees in tech to really fight or agitate internally in the company, or publicly as a whistleblower, instead of just exiting and finding something new? Do you think these shifts are possible, and what’s required for us to get there?

I: They’re definitely possible. But at the root of all of it is the American myth of individual mobility. The thing that keeps people in any industry from not working together with their fellow workers is thinking that they can be the next boss. 

People are so focused on being the next winner that they don’t understand that the system is set up for them to be losers for the rest of their lives. It doesn’t matter whether you make a million dollars as a senior software engineer or $100,000 as a junior software engineer or whatever else. You are fundamentally in the same rung as the line cook, as the janitor, as whoever else. You’re either the boss or you’re not. If you’re not the boss, even as a middle manager, you’re still a worker, and they will dispose of you the same as they’ll dispose of everyone else under you. 

Either we all work together towards getting our rights, or everyone is in this precarious position where they can be discarded. 

M: Was that something that you were always keenly aware of before your first job after graduation? Or was there kind of a pivotal moment in your career where that became extremely clear to you?

I: I have always sort of felt that way, I have understood intimately that you are either the person in charge or you’re not. And even many CEOs are not the person in charge. The current CEO of Amazon is not the person in charge — Jeff Bezos still is.

My politics and knowledge of what I personally have the ability to change has evolved, necessarily, as I’ve gotten more experienced, but I have never really had respect for people just because of their seniority or leadership positions, which has gotten me in trouble. Basically, my whole life, especially as a Black woman, and as a Black girl, in school and in jobs, I’d say, “No, actually, I’m not going to defer to you just because you sense that I should.” You need to help me understand what it is that we’re trying to do, because just saying, “I’m the boss” is not going to work for me. That is something that gets people in trouble, and it has a deterrent effect. No one wants to be the one who’s yelled at in a team meeting. I’ve been in that position because I just wouldn’t let something go. No one wants to be the one who’s pushed out of a role, and I’ve been in that position too. There are so many people who reached out and said that they experienced or knew about everything that I spoke about at Pinterest, and they still had been unwilling to say anything.

Education, Action, and the Future

M: In this issue of Kernel, we’re taking a look at the people, tools, and ideas that are bringing us to a more just and prosperous future. From your perspective, what do you see that is moving us in that direction?

I: A renewed unionization effort in the U.S. is really important. There was a time when workers’ movements were brutally crushed through the murder of activists, union leaders, and union members to repress worker power. That we’re witnessing a labor comeback because of the desperate conditions people are in again, a century later, is both encouraging and depressing, because with every uprising, there’s going to be brutal repression. We’re not there now, but I don’t think we’re too far off from that. This is something to watch out for in any way that we can.

M: Finally, do you have any calls to action for the folks who will be reading this in Kernel magazine? 

I: Figure out what your area of expertise or interest is and then do what you can to change or push for change from that position. Not everyone is going to be interested in or capable of moving legislation — you don’t have to be. People can make changes inside their companies, inside their organizations, inside their universities, and they are going to know what that change is most intimately because they are in the position to understand what’s going on. Who knows if in five years you’ll be in a position to even make the kind of change you’re deferring to the future? The key thing is to do something now. 

Reboot publishes free essays on tech, humanity, and power. We’ll be sending out a few more Kernel preview pieces this week; if you’d like to get them in your inbox, subscribe below ⚡️

Subscribe now

✍️ Community Response

by Alexa Jakob, electrical engineer and climate activist

“The purity stuff… it’s just not practical to me” is the most impactful line of this piece for me, and shows why an organizing approach based on identifying, gaining, and wielding power is important. Though Ozoma does not explicitly use the term, her approach is similar to power mapping, an organizing strategy that researches key decision makers, maps their connections, and identifies the levers they can be influenced by. By gaining a deep understanding of how systems work, coupled with her willingness to meet with big tech executives, Ozoma is able to deftly wield the levers of power.

As organizers, I sometimes wonder if we let the perfect be the enemy of the good, and I sometimes feel that we shy away from wielding power. For a student activist group, refusing a meeting with a university president may refuse the co-optation of the movement by administration, but it would also leave an opportunity for influence on the table.

This complexity exists all the way up to Congress - Alexandria Ocasio-Cortez, probably the most prominent left member of Congress, often talks about rejecting this framework. During the summer debate over the bipartisan infrastructure framework, which would have invested in roads and bridges but also in fossil fuel pipelines, she tweeted, “Very often I am told that raising concerns abt provisions that risk harm to communities like mine means I’m “letting the perfect be the enemy of the good.” That phrase often becomes a euphemism for “I’m willing to sacrifice your community for mine.” I reject that. We can all win.” And yet, her refusal did not alter the outcome - the bipartisan infrastructure framework passed unchanged.

An inside-outside approach is necessary. We need activists moving the Overton window, doing nonviolent direct action, holding rallies. We also need to be in every room where every decision is made. To work towards justice, one should worry less about keeping a pristine conscience by never making decisions, and instead do the difficult work of grappling with our responsibility to use our power to work towards justice. We won’t get personal absolution for our inevitable mistakes through this work, and we don’t need it. We just have to keep chipping away at systems of oppression.

Thanks for reading! We’d love to hear your own response to this conversation — drop a comment on this post or tag us in our thoughts on Twitter at @kernel_magazine.

Order Kernel Magazine

We’ll be back tomorrow with a reported piece on Ameelio, a tech nonprofit aiming to topple the prison communications industry.

Pulling the right levers and pushing the right buttons,

Emily & the Reboot team

Karen is the senior AI editor at the MIT Technology Review. She’s also an MIT Knight Science Journalism Fellow, wrote the newsletter The Algorithm, and co-produced the award-winning podcast In Machines We Trust. Prior to the Tech Review, she was a data scientist, an application engineer at the first startup to spin out of Alphabet’s X, and received her BS in Mechanical Engineering and minor in Energy Studies from MIT.

Karen’s commitment to rigor, empathy, and curiosity is reflected both in her approach to reporting as well as to her thinking on ethical tech careers. She also offers a peek behind the curtain at the impactful yet often controversial work of tech journalism, and includes advice for aspiring journalists and technologists alike.

Subscribe now

🗞 Peeling Back the PR

This event was facilitated by Reboot community member Archana Ahlawat for Reboot’s private community. This conversation has been edited for length and clarity.

Welcome Karen, and thank you for being here! To start with, how did you begin your career path in tech?

After college, I moved to San Francisco to work at startups. I was obsessed with how we could fight climate change, which is obviously a very long and laborious process, but I didn't see the venture capital model being the right vehicle for tackling longer term issues. 

I was at a company called flux.io, the first startup to spin out of Google X. Flux was trying to use technology to incentivize more sustainable architecture and urban development, so I thought it was the perfect alignment of technology with a problem that I wanted to solve. But Flux failed spectacularly. If you go on Google X’s website, it's scrubbed from ever existing. When it failed, it made me take a step back and think, Is the tech industry the right place for me? 

I had no idea what to do next. I thought, If I'm not going to be in the tech industry, what are spaces that are not as driven by a for-profit model? It came down to government, nonprofits, or journalism. But I didn't have a background in any of those worlds because I had been so focused on getting into the tech industry for so long.

How did you navigate your transition from engineering to journalism?

I first started exploring journalism because I love writing and it was something that I'd always been curious about.

In the Bay Area, there are two strong journalism schools, one at Stanford and one at UC Berkeley. I took day trips to each of them, sat in on classes, and talked with students there about what it means to be a journalist and how I could marry what I want to do with a career in this world.

And I just really fell in love with journalism. I loved the classes, I loved speaking with the students, I loved the way they were thinking about their role in society, and I was excited about being able to work for mission-driven organizations. In journalism, there is obviously a looming need to make money, but as a journalist, you're split from that. There's a strong divide between the business side and the editorial side. 

I didn't know any other journalists, so I went to the MIT alumni database and searched for every person that had ever done journalism. There were around 50 people in a span of 70 years, so I reached out to all of them. Maybe 15 people responded, so I got on calls to ask how they made the transition from studying engineering.

Two paths emerged. You either go to grad school or you keep writing, publishing, and trying to get an internship until you make it. I was like, Great, I don't have money to go to grad school, so I'll just write until I make it. I learned to write an article in an edX class, shopped that one article around, landed an unpaid internship, went from there to a fellowship, went from there to a staff job, and that's how I made my transition.

“I’m not afraid of reading technical papers or talking to startup founders or engineers about how their technology works. So I started differentiating myself by digging into my technical background.”

How did you figure out the unique perspective you bring to writing about the tech industry?

I originally wanted to be an environmental reporter, so my first unpaid internship was at the Sierra Club magazine. 

But people don't hire environmental reporters. In 2016, everyone wanted tech reporters. It was when the techlash was starting, and there was a lot of interest in the tech industry, its culture, and new technologies like AI. Because of my resume, whenever I applied for jobs, they would ask me, “Do you want to be on the tech desk?” 

I took my first tech writing job pretty begrudgingly, but realized that I could make this into a strength. I’m not afraid of reading technical papers or talking to startup founders or engineers about how their technology works. So I started differentiating myself by digging into my technical background. Also, the reality of journalism is that you have to draw upon your network, and mine was in tech, so it was a lot faster for me to identify who I needed to talk to and get conversations.

But as I continued reporting, I realized that there was also a fundamental misunderstanding of the culture in the tech industry. Some of the smartest and kindest people I know work in tech. They have amazing intentions and then work for companies that do not benefit society in the ways that they espouse. I was fascinated by what was going wrong in this process. For people who haven’t been close to those who work in tech, there's a perception: What are these people doing? Why is this stuff always messed up? Maybe it's not about the people, it's about the systems that they're in. 

I also started thinking, What do I need to tell my friends in the tech industry about the world outside? A lot of people in tech get really insular. You're living in San Francisco, interacting with other tech people, and don't think about other perspectives. I wanted to write stories that help people who might be drinking the Kool Aid realize that there's other things they should be considering when they're building technology. I wanted more bridge-building and conversations on both sides of the aisle.

It does seem like there are fewer such gaps in tech reporting now — many journalists have a more holistic grasp of the industry. Where do you see that there's still work to be done? 

There are a lot of parallels with the tech industry. In tech, we're seeing younger technologists question whether technology is truly neutral, like there's bias that's embedded within these technologies. 

In journalism, the same transition happened. Old school journalists considered themselves to be very objective. They would act like they don't have a position.

I’d consider myself as part of a younger generation of journalists throwing that out the window. There's an argument behind the things that I write — I'm always very clear about what my position is. I don't think that I'm neutral, and I don't think that I should pretend that I am. But I strive to be fair and honor a multiplicity of perspectives. 

“I try very much to respect the people I talk to, even if I'm critiquing the systems or the company that they're part of.”

There's a lot of PR and narrative-building coming from the tech industry, yet you also seem to have a strong sense of what's going on within the academic sphere. How do you manage that?

It's really just a race to keep on top of all the AI research that comes out. 

When I started covering AI, I would read five or ten papers a week, then get on calls with researchers at the forefront of their fields. Once I was able to build a core understanding of AI’s limitations and capabilities, it was easy to see a startup saying “We're going to solve AGI in two years” and think, That's an interesting claim that most of the research community does not agree with.

Over the years, I’ve developed a lot of good friendships with AI researchers, so I'm always checking in to ask: What's new? What do I need to know about? How are things shifting? All the best tech reporters do that. They stay close to the research and have their own understanding of it before they start talking with companies or PR people that have the financial incentive to aggrandize what they're working on.

It can be exhausting. There were times when I would go on vacation for a week and feel like I missed an entire shift of the industry.

How do you maintain relationships with the tech industry despite some of their skepticism of the media?

When I write something that brings PR jargon down to reality, I've never had someone get upset. They respect that my articles are technically grounded. Sometimes, in a call with a company representative, I'll ask about the technical details and it’ll catch them off guard — I understand more than they thought I did. Then, they’ll tell me a bit more about the honest capabilities of their system, and that's how I gain respect and trust with those individuals. 

In cases where I've had a more contentious story-writing process, most recently with Facebook, I try very much to respect the people, even if I'm critiquing the systems or the company that they're part of. It's a hard balance to strike. I don't know that I always do it correctly, but at the end of the day, I still feel comfortable calling up the individual that I spoke to and feeling like I honored them as a person. Even if I am very harsh on the things they did, I try to be clear that their actions don't necessarily make them evil or bad.

How do you research questions that require access to corporate data?

I try to be realistic about what I have access to at the start of a story. I wrote a piece about how Facebook Instant Articles is funding misinformation in the Global South, and that was based on open source data and the CrowdTangle API. I scoped the story based on exactly what I knew was in CrowdTangle.

So one limitation of reporting on companies from the outside is that you have to work with what you're given.

How do you sift through everything going on in the tech industry to choose what to focus on?

Something that really surprised me when I came into journalism was how subjective everything is. It’s just based on what I'm interested in. I spend a lot of my time absorbing information, having calls, reading papers, looking at the news, and just keeping a pulse on conversations. When things grab my attention, I’ll start to chase.

I always try to think back to my ultimate goal. I gravitate towards stories that reveal how tech culture shapes technology and how technology impacts people. I also consider: If I didn't write the story, would someone else be able to? Is there a particular thing I have that would help me write the story in a way I think is valuable, whether that's access to particular people or the technical ability to explain a complicated concept?

“There's an argument behind the things that I write — I'm always very clear about what my position is. I don't think that I'm neutral, and I don't think that I should pretend that I am. But I strive to be fair.”

How do you think the training of engineers impacts the culture of the tech industry?

One of the most obvious things in my undergrad was that I never had to take a single ethics class. At MIT, we had a humanities and social sciences requirement, but the culture within the student body was that these classes were a joke. That directly feeds into some of the challenges we see in the tech industry where people don't value certain types of expertise that are core to building technologies that are better for humanity.

The other thing that bothered me when I was at MIT was the career fairs. It was this huge thing with hundreds of companies. You're trying to figure out how to make a career after four grueling years of education: what's going to pay you, what other people see as exciting and prestigious. When you walk into the fair, tech companies like Snapchat have the largest crowds. There were never any social impact companies or opportunities, so you just assume that they don't exist. 

If I knew then that you could be a tech-interested person not directly in the Silicon Valley startup industry, I might have tried a whole assortment of other things. But they don't show up in these signaling venues, so students don't think about those opportunities.

Do you think if you had graduated at this time, you would find more values alignment in the tech industry?

No, I don’t think so. I have friends who are excited about using data science for social impact, but also really want to do technically challenging work. There's only so many companies that you can do that at, so they move from like Facebook to Twitter to whatever, and ultimately, those aren't companies I think I would want to stay at.

Instead of journalism, I probably would have ended up at a research institute or think tank at the intersection of tech and policy — things I didn't know existed in college. Maybe I would have stayed, or who knows, maybe I would have become a journalist anyway.

🌀 microdoses

• For more perspectives on the social impact of journalism (and people who read more than I ever will), listen to Ezra Klein’s conversation with the 80,000 Hours podcast.

• Worried about catching Omicron? Read this very scientific and factual list of symptoms, from aspirational Zillow shopping to “longing for the idea, if not the reality, of fondue.”

• What happens when scientific consensus collides with political pluralism? Here’s Hannah Arendt’s take on why truth and objectivity are much more complicated than they seem.

💝 closing note

We're prepping for Valentine’s day…. Tell us a tiny (tech) love story (100-200 words)! We’ll allow anonymous/pseudonymous submissions. Send them in here!!!

If you hate romance, here’s our regular pitch form.

Aligning our values,

Reboot team

Karen Hao is the senior AI editor at the MIT Tech Review. We’ve learned a ton from her intensive reporting over the past several years, from breaking down the LLMs paper that got Timnit Gebru fired to her recent work on “How Facebook and Google fund global disinformation.” You can check out her website

Read more

I’m super excited to share today’s interview with Hal Triedman (he/him), a Privacy Engineer working on transparency and machine learning at the Wikimedia Foundation.

🌐 where algorithmic transparency meets community

This interview has been edited for length and clarity.

What do you work on at Wikimedia? 

If you were to abstract everything that I do into a single phrase, it would be algorithmic transparency. That encompasses making model cards and trying to make sure that people are aware of and understanding the models that are affecting their experience on the platform, but it also includes things like creating transparency in the sense of releasing more data. We don't collect a lot of data — the privacy policy explicitly states that the Wikimedia Foundation minimizes the amount of data collected — but we're a really big website. Wikipedia gets somewhere on the order of 13 billion pageviews per month, which puts us in the top five most looked at sites in the world. We're the only nonprofit in the top five, and because of that, there’s a unique responsibility to shine some light into how the Internet is actually working. 

What kinds of algorithms is Wikipedia using? If I were interacting with the site, where would I be actually interfacing with an algorithm? 

We have 150-200 models spanning 30-40 languages; for the most part, they’re designed explicitly as human in the loop systems. Only in very rare cases will a decision be made solely based on a model prediction without a human sign-off. 

One example of how models are used is edit review. Let's say somebody goes to the Wikipedia page for OpenAI and ― anyone can edit Wikipedia ― let’s say they really don't like what OpenAI is doing. They hit the edit button and they type something vulgar. That’s not a good faith edit, and we have a model to check for that: saying this has a 75% chance of being a bad faith edit. This model is about trying to make the life of editors who are trying to sign off on edits a lot easier by prioritizing the edits that are potentially going to be harmful to the community and the quality of information. 

These aren’t super advanced models, to be clear, more like a random forest classifier than GPT-3. 

Other examples of where we use algorithms: classifying the quality of articles ranging from stubs to fully fleshed out articles which we call Featured Articles; those are the articles we could feature on the home page because they're incredibly high quality. There’s also topic classification ― sports, biochemistry, philosophy. Across languages the models are often quite similar in architecture since the tasks are quite similar, and because we’re looking at features of individual edits rather than features of big chunks of text.

From an organizational perspective, how and where did this push for transparency start?

We're not a very large organization, probably something like 500 employees; of those 500, maybe 300 people are technologists. So it's a small staff working on a large ship. A lot of these questions about transparency have been in the air from people at the very top ― director level, VP level, all the way down to individual contributors for years now; these conversations were definitely happening before I got there in April 2021. It’s a pretty institutional concern. Wikipedia is in an interesting position as an incredibly large presence on the web; and people who engage with us, whether they’re donors or editors or contributors, generally do so with the assumption that the wiki process will eventually converge on some semblance of “good information,” of neutrality. Whether that’s a fair assumption to make is probably outside of my purview ― but there’s a lot at stake, so we want to show everyone our cards, do everything open source, and in theory, make it possible for everyone to understand or for interested outsiders to check our work to make sure that we're doing things the right way.

Can you talk a little more about the scale of Wikipedia (and its contributors), and how that affects Wikimedia policies? 

Let’s zoom out for a second. There are hundreds of millions ― maybe billions ― of individual people who are seeing a Wikipedia page on a regular basis; within that haystack, there are probably a couple hundred thousand people who are editors, who are really interested in their specific subject area ― military history or organic chemistry or scientific papers ― and they keep their little corner of Wikipedia up to date as best they can. Within that group of editors, there is also a smaller group, let’s call them super editors, who are more oriented towards the organization; that’s probably a couple of thousands people. So have an outspoken, vibrant, intense community of a couple thousand people who are poring over everything that we [Wikimedia Foundation] do and giving their input on our decisions; there are also researchers who are looking at how our decisions are affecting, for example, gender parity among new article creators or article content. So there’s a built-in urge to transparency as a function of the fact that we as an organization rely so heavily on this external community.

Still, there are sometimes also cases where that runs into privacy issues ― we don't want to accidentally release user data, or break anonymity, or, more concretely, cause a government to take adverse action against an editor. So we do have some conflicting urges, between the urge for radical transparency and the urge to make sure that all of our users are safe. 

Wikipedia thrives off a robust, decentralized community of contributors and decisionmakers, and yet from a technical perspective is a collection of static pages with links to one another. Do you have any thoughts about what Wikipedia might tell us about Web3, and how the ecosystems there might develop? 

I'm by no means a Web3 expert, and Wikipedia is not perfect, but I think that overall, out of all the technology that came out of the early 2000s, I think that Wikipedia has stood the test of time as a pretty unequivocally good force ― especially as we now exist in this world of fractured epistemologies, and untruths, and the mistrust that pervades the social fabric in America but also all over the world. Wikipedia’s not going to solve every problem but it’s a pretty good thing that came from that era. 

The interesting thing about Wikimedia [as a technical organization] is, from 2001 until around 2014, 2015, it was run by very few people. They hacked together some code and got some servers up, and held it together for more than a decade with shoestring and duct tape. What we see on Wikipedia today is really about the community and the mission much more than the technology. 

A take away that I've increasingly had ― as someone who's neither a techno-optimist nor a techno-pessimist but more of a techno-realist ― is that the technology is not going to be the thing that makes it work. It’s less about the technical features of a platform ― immutability, provenance, et cetera ― and more about the community, the norms, the people. In other words, the squishy stuff around the edges. 

What have you been reading recently?

Lots of history and fiction — right now I’m reading Devil in the White City by Erik Larson, Empress Dowager Cixi by Jung Chang, and Midnight’s Children by Salman Rushdie. I just finished The Sympathizer by Viet Thanh Nguyen.

Find more of Hal on Twitter, on Goodreads, or at his website.

🌀 microdoses

• 🐇 If you want to go down a rabbithole about Wikipedia, here are a bunch of links to get you started. (For a standard Wikipedia rabbithole, the homepage has plenty to offer. Did you know that 140 years ago today, a meteorite landed near an Ukranian village? Also, happy World Toilet Day.)

• 📜 RIP ConstitutionDAO… but it’s not too late to check out this explainer in VICE, written by our very own Lucas Gelfond.

💝 closing note

Last week, we published a proposal for a corporate wealth tax in A Tax on Tech-quity. Reader Chris Beiser comments: “One adverse consequence of taxing equity is that it advantages fast-growing companies like startups, which have only existed for a short time, at a cost to businesses that grow slower, in more prosaic parts of the economy. This would push more capital towards "those who own the technology companies building these advances” — the opposite of what this proposal claims to seek.” His comment is worth reading in full — check it out below.

Reboot

⚡️ A Tax on Tech-quity

This week, we’re sharing something a bit different: a tax policy proposal from AI policy aficionado, Kernel Magazine author, and Reboot community member Chris Painter. While this essay seems more economics than technology, we think it gets to the heart of debates over how to ensure technological advancement benefits everyone — not just the biggest corpo…

Read more

4 years ago · 4 likes · 1 comment · Chris Painter

What are your Wikipedia core memories? Did you also get yelled at by your teachers? Are you a contributor or editor for Wikipedia, and what’s your experience been like? Reply to this email or comment below — we’d love to hear from you.

See you on the Wikipedia edit logs,

Reboot team

Today’s interview is with scientist and longtime Reboot editorial lead Ben Wolfson. Ben earned his PhD in Molecular Medicine from the University of Maryland Baltimore, followed by a postdoc in cancer immunotherapy at the National Cancer Institute. Currently, he serves as the Assistant Editor-in-Chief of The Journal of Science Policy and Governance and as a Science and Technology Policy Fellow in Washington D.C.

🔬 pursuing the endless frontier

This interview has been edited for length and clarity.

What have you been doing for the past few years, and what are you up to now? 

I got my PhD in cancer biology in 2018; since then, I’ve been a postdoctoral researcher in cancer immunology, and now I'm about to start as a Science and Technology Policy Fellow. I’ll hopefully be working on policy governing what research gets funded and how science works in the federal government. 

Why did you decide to make the switch from wet lab research to a policy role? 

For me, policy has been an ongoing, slow-burning interest. There’s a lot of people who can do good science and who really want to do good science. But to enable good science and technology, you need good politicians and good policies that promote innovative research and guidelines for doing so in equitable and ethical ways. 

Policy is a field that needs scientific expertise. The purpose of the fellowship I’m doing is to bring scientists into government, specifically into roles where they can offer their scientific expertise, while simultaneously getting trained on how to change and contribute to policy. 

Right now, scientists have to deal with the fight for funding and the fight to do research that is unimpeded by unnecessary policies. Cancer research is privileged in that everybody understands the need for it, so there's always lots of funding, but this isn’t always the case. I’m thinking specifically of fetal tissue research bans, which are often pushed by anti-abortion activists. Some people might think that using fetal tissue is unethical, but it’s a critical resource for a lot of important work in medicine.

What do you think about the nature of research when it’s privately versus publicly funded?

The entire structure of the government’s scientific research funding right now is built on a document that was written in 1945 by Vannevar Bush called the “Science, the Endless Frontier.” It argued that as World War II ended, the U.S. government should continue to fund science, and especially basic research, at wartime levels in order to promote growth. 

Basic research is super important for the development of scientific applications, but it isn't as well-funded by private industry because companies can't necessarily get something profitable out of it immediately. That's the dogma for why federal research is important. 

However, federal research funding has been decreasing or stagnating for the past several decades — which means that in many important fields, funding has been taken over by the private sector. We're starting to see incredibly wealthy companies devote billions of dollars into research at unprecedented levels. 

Alphabet has the biotech company Calico Labs, but it seems like they’re struggling a bit — they haven’t done a whole lot publicly, at least. Other lab science that tech has funded seems to be focused on their founders’ priorities, which often happen to be life extension. Then, there’s Apple’s investment in personal health tech via Apple Watch, but it’s all based on what might be profitable, and that’s not necessarily the reason you do something in federally funded science. 

Is science policy just a question of funding then? Or are there other limits beyond finances?

I’m interested in more than just funding. That’s another rapidly growing field: What promotes innovation? How can we speed it up? These questions were jumpstarted by Tyler Cowen and Patrick Collison a few years back.

One of the problems right now is that we can throw a lot of money at research, but we’re not so good at connecting work in disparate fields together to figure out what’s valuable. For example, I think we're in the very early stages of seeing what machine learning can contribute to biological science. Google just announced a month or two ago that they’d cracked protein folding, which is a huge deal if true. We’ll probably start seeing more stuff like that. 

Are you worried about the potential for research itself to be privatized? For example, would Google ever refuse to release their protein folding work? 

It's certainly a scary possibility. So far, the Supreme Court has made it nearly impossible to patent genes and some similar things. But a company can still restrict information or choose not to publish findings that they think they will be able to make profit from eventually. 

That being said, there’s a very strong culture in science about the freedom of information and sharing results. So if that happened, there would be blowback. At the same time, I wouldn't be surprised if in the case of blowback, the scientists speaking out were fired, and then the company continued to keep its research secret. 

The culture of Western science has always been one of “freedom,” but in reality, it’s sometimes been more of what we say than what we do. 

I read a great book called Freedom’s Laboratory about how the culture of scientific freedom in the West was used as propaganda in the Cold War against the Soviet Union. Even though it started as propaganda, that culture has become self-fulfilling; most scientists very dearly believe in open science. Even right now, we’re seeing pushback against publishers who charge exorbitant fees for manuscripts. That’s a good direction, but that’s only for work that’s been published — and companies may not have an interest in publishing their work. 

You mentioned that the push to fund basic research has roots in World War II and the Cold War. What do you think about that framing? Do you think there’s a way to foster innovation without the explicit goal of winning a war? 

Military framing is a very easy way to sell research or to sell funding. We're seeing that right now with the USICA, which is a large research funding bill that is being framed specifically as a way to combat China, in a pseudo-Cold-War mentality.

I have complicated feelings about it. The metaphor of war is not necessarily the most productive or the most ethical, and of course, it would be nice to secure funding by saying, “This will be good for everybody.” But if wartime framing helps it get done — like if we have to go to war against global warming — then maybe the ends will justify the means. Nixon declared a war on cancer, for example, which increased cancer funding substantially. We haven’t cured cancer, but we’ve definitely made progress since then. 

It seems like the only way politicians know how to describe research goals is war on or moonshot for — we had a Cancer Moonshot under Obama — but the strategy sells. 

Any parting thoughts for readers considering science or tech policy careers? 

I got into science because I love science fiction and thinking about the future. That’s the same reason I’m getting into science policy: I see what’s possible, and I want to enable it. That can just as easily be done on both sides of a lab bench. 

The same is true for other fields — if you have a vision about what science and technology should look like, you can push for that from no matter where you are. 

Find more of Ben on Twitter or his website.

Subscribe now

🌀 microdoses

• 🚀 DARPA is the federal agency behind breakthrough technologies from the laser to the internet. Read this insane deep dive into how and it worked so well.

• 📚 This New Yorker story describes how Overdrive’s consolidation chokes public library budgets.

• ✍️ All writing is collectively created: an amalgam of one person’s perspective with the innumerable bits of knowledge that influenced their thinking. In this essay, Sari Azout and Jad Esber explore how new business models might recognize the cooperative nature of creative work.

• 🔎 Satisfy your parasocial fantasies with this hilarious look at NYT tech reporter Mike Isaac’s Google search history.

💝 closing note

Someone in the Reboot community recently asked for written pieces of “tech satire”:

• Pat: I read “How to Become an Intellectual in Silicon Valley” a few months ago.

• Lucas: What Tech Calls Thinking, kinda!

• Jasmine: Uncanny Valley is near-satirical, except for the part where it’s real.

(But seriously, send us some satire — we could use a laugh.)

Research on,

Jasmine & Reboot team

I’m thrilled to publish this interview with Jason Prado, The Drivers Coop’s Chief Technologist. It’s an inspiring example of how engineers can “go where workers are and organize with them” while leveraging the scale that technology offers.

Subscribe now

🚕 putting workers in the driver’s seat ft. jason prado

This interview has been edited for length and clarity.

How did you figure out what you wanted to do, and what are you doing now?

I’m currently the Chief Technologist or Head of Product at The Drivers Cooperative, a driver-owned ridesharing service in New York. 

I studied Computer Science at Stanford from 2004 to 2008. I always knew I was going to major in CS because I had been playing with computers since I was five years old. I’m also the first person in my family to go to college or have a professional career.

In college, I interned at various companies and startups in the Bay Area. After graduating, I worked at Microsoft, Google, and Facebook. I was enamored with the tech industry and totally bought into the Californian Ideology. I worked at Facebook for about seven years until 2020. Over those years, the world changed, my politics changed, and so did my view of Facebook. 

The 2016 election was a big wake-up call. I realized that maybe my love for Silicon Valley ethos was a bit unfounded. I started looking for other ways to change that I believed better reflected reality.

I joined the DSA the day after the 2016 election and have been active for the past five years. That led to getting connected with other tech workers interested in labor issues and union organizers in the tech industry. I started working on labor campaigns on Facebook campus for security officers, cafeteria workers, and my coworkers.

During my last two years at Facebook, I worked on the cryptocurrency project now called Diem. While I thought some of the goals of the project were worthwhile, I didn’t really believe in what Facebook was doing overall. The goals of Diem seemed so lofty compared to Facebook’s misinformation campaigns, monopolistic practices, and impact.

Eventually, I left. I joined Bernie Sanders’s presidential campaign as a data engineer before Super Tuesday. After that, I spent lockdown doing research on cooperatives and collaborative economy. And five months ago, I found The Drivers Cooperative. I joined the organization as the first and still only full-time technologist.

Why are you excited about The Drivers Cooperative?

A platform cooperative is like Facebook, but owned by all of its users and democratically governed. 

I found The Drivers Cooperative through Start.coop, a venture fund and an accelerator for cooperatively owned businesses. I became an investor in Start.coop and gave TDC some advice on their tech decisions as they are one of our portfolio companies. 

What drew me to the organization was people coming together to solve their own problems. Thousands of drivers in New York City were sick of Uber and the bad deal they were getting. They started asking: “What if we do this ourselves? We have the actual labor here. Can we build our own rideshare service? Can we make tech work for us instead of us working for tech?”

I'm very skeptical of tech guys who are like “I know what these people need, it's an app!” The Drivers Cooperative had an opposite approach. These were very dedicated, motivated people, who knew a lot about rideshare and managing logistics. They literally just didn't have a tech person. That's why I wanted to join. 

After joining The Drivers Coop, what was most surprising to you about what workers and drivers need?

In New York, most rideshare drivers are professionals. They're very good at their job and take it very seriously. Yet they're all scared of being deactivated by the Uber algorithm. Their livelihoods can be cut off from them and they can't even talk to a human about it. It sucks to work for someone and not be in control of your destiny. 

The Drivers Coop has regulations and operating procedures, because we're a business, but the ultimate say for these policy decisions is made by the Driver Board, which is a board of drivers elected from among the driver membership. 

What are the key challenges for the organization right now?

It's hard to compete against Uber and Lyft with their great applications, so tightening our operations and building tech are our biggest challenges. To do that, we have to raise capital, like our successful crowdfunding campaign from two weeks ago. Before the campaign launched, I didn’t believe it was going to succeed, but now it's starting to feel a little bit real. I'm now spending all my time hiring people, and the quality of candidates has exceeded my expectations. There's a light at the end of the tunnel, but it's going to be a difficult six to nine months to transition.

From a technical perspective, you’ve mentioned that your app just isn't as good as Uber. How much does feature parity matter, and how do you win on feature parity when funding is so constrained?

Winning looks different for us than it does for Uber and Lyft. Uber raised tons of venture capital and an absurd valuation. To justify the valuation, they have to win a monopoly. Our strategy has to be different. If we want to employ 10,000 people, help them provide for their families, and be sustainable, we don't have to fight tooth and nail to destroy all of our competitors. Our job is to prove that this works, and it can scale. 

But there has to be some other intervention to beat Uber. I think that if you are a local government, you should make it very difficult for Uber to operate. Uber is an exploitative player that takes wealth out of your community and gives it to the people who already have wealth outside of your community — you shouldn't want that in your neighborhood. On the other hand, you should make it easy for cooperatives to operate; you should give them funding; you should give them friendly policies. That's the long term play.

Have you seen interest from drivers in moving into office and operational work, and is The Drivers Coop building that career pathway as a possibility? 

There are a lot of tensions here because we have to operate and build stuff. I am stretched pretty thin and our tech stack is going to be stretched pretty thin forever. Uber always took the more technically challenging decisions because they had access to infinite capital. We're not in that situation, so we just have to get business done. At the same time, if we just do a crappier version of Uber and don't empower anyone in the process, then we're not going to succeed.

We’re working on programs to rotate drivers through roles in our offices. Our success will depend on how closely aligned the driver-owners and the staff stay, so this is a priority for us.

In our job listings, we say we want people with the relevant skills and prefer people who have experience in gig labor platforms. That being said, the reality is we also need people who have worked in Big Tech — there's knowledge that you only have if you worked at FAANG. We need that to build things that scale. So it's really, really difficult. 

What are your hopes and dreams for what The Drivers Coop can become?

I've been thinking about an API that describes gig work at a semantic level. You would have different players competing at every level of the stack, like your local food delivery cooperative could have multiple front end apps. It’s very early, but we’re having conversations with other cooperatives about what interoperability might look like. If we figure out the one true API for gig work, then we can start to muscle in on Uber’s territory the way Mozilla tried to challenge the monopoly browser platforms. 

Interoperability is a hard path compared to one guy on top telling everybody what to do. Maybe everyone who works on cooperative platforms could come together to challenge that. That's something that you can't reason about abstractly, you just have to try.

What do you think about DAOs being used to form coop-like structures?

I talk to VCs whose first question about The Drivers Coop is “Is this a DAO? I'll invest in a DAO.” It’s like, come on. Our driver base is over 90 percent immigrants. We have some tech-savvy drivers for sure but it’s not the norm. Most of our problems are very operational, like can we get some documents uploaded from a driver and get them driving.

That being said, our platform has to scale. We want to scale it to other verticals and scale all kinds of management processes. I think DAOs are kind of cool, and scaling to two million people is something that they might enable someday — to run this congress of every gig worker on earth and build the one true labor platform of the future. 

There's a lot of potential in blockchain, but we're not guaranteed to have emancipation through such technologies, so I approach it with a reasonable amount of skepticism. 

What other coops or alternative structures would you want to direct more attention towards?

I would keep an eye on Start.coop, who makes it through their accelerator, and watch them for the next few years.

A lot of young technologists are balancing the desire to work on values-aligned projects like coops versus learning skills at FAANG companies. There isn’t any one size fits all advice, but what would you recommend for folks in this position?

It depends on what you're looking for. Definitely prioritize your financial situation — pay off your student loans, support your family.

Labor organizer Jane McAlevey wrote this book called No Shortcuts, which is about strategy for labor and progressive organizing. Somebody asked her what they should do if they are a young student and want to support the movement. Her advice is to get really, really good at your job. The labor movement doesn't just need more staffers. What it needs is more people embedded at workplaces and more people with the right mindset as our leaders. The best organizers are the best workers, so get good at your job and become respected in it, so that when you lead a workplace struggle, people will follow.

For people who are straight out of undergrad and thinking about whether to join something impact-motivated or to get in two years at Google, I don't know. I'm a little biased, but having Google on my resume has carried me a long way. When you tell people afterwards that you're working in a social impact startup, they’ll take you seriously.

If you want to prevent values drift, just don't get too comfortable. I’m still friends with folks from Google, and we focused on our careers together for a long time. But my community has changed over time — it is more socialists and organizers and critics of tech, which keeps me honest.

Subscribe now

Find more of Jason on his Twitter or Substack.

🌀 microdoses

• 💼 The Drivers Coop is hiring software engineers and data engineers! You can also support them by investing.

• 🖋 The CIA funded literary magazines as part of the cultural war against communism, bringing up bigger questions about the role funding plays in “radical” art and media.

• 🏫 Nadia Eghbal deep dives into why Arizona State University might be a novel future for accessible and effective higher education. (Way more interesting than it sounds.)

• ✊ Jacobin interviews Reboot community member and Times Tech Guild union organizer Angie Kim: “It’s not our job as organizers to insist that everyone agrees with us. Our work is to consider people’s concerns, work with them, and to organize around the overall concerns of the entire workforce.”

Steering ahead,

Jasmine & Reboot team

Jessica currently works on safety operations at OpenAI, where she thinks about how information travels through internet ecosystems and the potential of human-computer collaboration, and she grew and moderates the Facebook community Subtle Asian Dating. She's also tinkering with form as a 2021 Interdisciplinary Writer’s Lab fellow with Kearny Street Workshop and writing poems for strangers as a part of the 2021 Pride Poets cohort.

🛠️ Safekeeping the Internet ft. Jessica Zhou

This interview has been edited for length and clarity.

What do you work on at OpenAI?

At OpenAI, we see lots of applications built with GPT-3 for copywriting, customer support, product catalog applications, as well as creativity tools for writing.

I started off evaluating these applications against internal and external policies like the use-case guidelines, safety best practices, and terms of use. I've also updated our policies to reflect newly surfaced safety risks. I'm now focused on monitoring to better understand emerging themes in how people use the GPT-3 API, and I build reports and classifier training efforts to proactively facilitate safe application development. Finally, we support developers in putting policies into practice within their applications. There's a lot of humanness involved in reaching out to someone and talking them through our reasoning.

We aren’t directly moderating content, but we do want people to moderate their use of the platform, so it's almost like moderating someone else’s moderation.

Is there anything you were surprised by at OpenAI?

This isn't specific to OpenAI, but I’m surprised by the various ways "the market" exerts pressure on organizations to be financially sustainable to the point they have to compromise on mission, integrity, or effectiveness.

For example, during my first week into an editorial internship at the LA Times in 2016, I watched a bunch of people get laid off because of the great false-alarmed “Pivot to Video,” which eventually cost hundreds of journalists their jobs and reflected a number of deeply-rooted problems in the industry, like the volatility of digital advertising, over-dependence on aggregator platforms like Facebook, and out-of-touch business executives.

Language model APIs are a different domain, but what persists is the push-and-pull between business pressure and what might be best for society. For example, something that Brian Christian highlighted in his Reboot talk was the tension between research cross-collaboration for altruistic societal ends versus the business interest of each entity to stay siloed from one another.

In my head, I have all these ideals of how things should be, but I realize the process of actualizing ideals means having to make decisions about what to prioritize each point in time. I’m learning to accept that these decisions aren’t so much compromise as they are intermediary steps that we take to work towards an ideal.

I was also interested in your work moderating Subtle Asian Dating (SAD) — that's such an interesting context for thinking about online communities and the performance of the self.

SAD blew up over Thanksgiving break; we went from 1,000 to 100,000 members super quickly. At the beginning, we had a single intention — that this should feel fun and different from our other experiences with dating. During this exponential growth period, we had to very quickly put out working moderation policy. Even if it was hard to enforce in practice, it’s critical to say that we don't tolerate colorism, sexism, homophobia, transphobia, or other things that attack people's humanity. Then we got into really weedy questions; for example, how do you even know if someone is from the Asian diaspora? We wanted to be very porous in allowing as many people to be here as possible.

The work my team is doing at at SAD is unpaid even though it plays such a critical role. I’m certain we’ve generated a ton of money for Facebook; people kept telling me that SAD and the related constellation of groups were the only reasons they kept using Facebook. Moderation for SAD feels like a labor of love, but it’s surprising how so many hours of labor go unpaid by one of the biggest companies in the world. This gets at a tricky ongoing question between platforms and its creators and contributors: VCs seem to love "network effects,” yet so much merit is heaped upon those who build a platform. That has big landlord energy to me.

Another thing that was surprising from SAD was recognizing that we didn't want to trademark or be possessive of a concept that resonated with a ton of people within the Asian diaspora. People hosted independent SAD-themed events and started spinoff local groups. I met one of these groups in Austin, and it was such a heartwarming moment to see all the new friendships that sprouted. Not everything has to happen under our umbrella, and we were a part of growing something much bigger than ourselves.

Do you see any parallels between your work at OpenAI and at SAD?

It's really cool to be in a role at OpenAI where I'm paid to do the same things that I did with SAD — figuring out how to navigate potential conflicts between community members or customers and our organization’s policies. Being paid for moderation-esque work is a nice feeling, but the unpaid stuff that I was doing with SAD was still materially important from a humanist, interpersonal perspective.

Additionally, the interplay between the internal and external in shaping any entity’s values is interesting — the feedback loop of interacting with the outside world to figure out who you are at the core. OpenAI's moderation policies evolved based on continuous interactions with developers, and SAD's moderation policies were based on interactions with group members. This reminds me of how I’m stumbling through the process of figuring out who I am: trying out things, spending time around people, negotiating what is important to me, adjusting to those new values, and repeating.

Any hopes for the future of social tech?

I see Facebook as a public space or a phonebook where you can stumble upon someone, but I'm excited for whatever becomes the go-to digital equivalents for casual friend hangouts. The act of creating products or companies is a form of building for a future you want to see, and so many other kinds of imagination should have the chance to flourish too.

I’m inspired by Eternal as well as Somewhere Good for this reason. We deserve platforms that are tailored after us alongside big platforms which aggregate a broad public  There are so many people in this world, and just a small subset of people building socio-technical infrastructure for our future.

Finally, Facebook has enabled people around the world to find one another in ways we couldn’t have previously imagined — I know how much my life has changed because of it. But now that we’re here, I am also excited to see where people go after they’ve found one another.

Find more of Jessica around the web: for media/computing interests, go here; for art/literary interests, go here; see what’s bouncing around her brain on Are.na, and say hi on Twitter.

🌀 microdoses

• 🤖 Recent research from OpenAI shows how language models can be trained on datasets curated for specific values, such as the ability to understand that beauty is subjective.

• 🔊 On a higher level, listen to this conversation between Open AI CEO Sam Altman and Ezra Klein. They talk not just about AI's technological future, but who's funding and governing it, and what that means for the balance of power.

• 🔏 5000 words on "How the New York Times Crossword Became Too Big to Fail."

• ⚖️ Speaking of too big to fail, Matt Stoller explains the significance of Lina Khan chairing the FTC (TL;DR: antitrust is back, baby).

💝 closing note

Ted Chiang is one of the Reboot’s community’s most loved authors. Here are a few of our favorite short stories:

• Jasmine: "The Truth of Fact, The Truth of Feeling" — I love anything about memory, and Chiang mixes in some great McLuhan-style reflection on written versus oral traditions.

• Pearl: "Anxiety is the Dizziness of Freedom"

• Laura: "Seventy-Two Letters," I'm amazed at the amount of Borges references he has in that story; and "The Tower of Babylon," because it was the first one that I read and stays in my heart forever.

P.S. If you’re a young person looking for a community thinking critically about tech, humanity, and power, Reboot runs a private Discord — more on that here.

Toward a better Internet,

— Jasmine & Reboot team

To find out, we talked to Angie Kim, an engineer on The New York Times's Machine Learning Platform team and a member of the organizing committee for The Times Tech Guild (check out their site for background on the movement). She's also a recent graduate of Brown University and an avid crocheter.

🛠️ Why I'm Unionizing ft. Angie Kim

This interview has been edited for length and clarity.

How did you get involved in organizing for The Times Tech Guild, especially so soon after you joined the team?

The first month I joined The New York Times was super chaotic. We were rushing to hit a deadline, so we were spending all our time on Google Meets and working late hours. This resulted in the crumbling of team morale and goodwill, and eventually caused a total reorg.

As I was venting to a mentor and friend at The Times, she told me about a group of tech workers who are "working on improving our workplace." I immediately recognized this as a signal of workplace organizing: back in May, I had attended a panel hosted by the NYC-DSA Tech Action Working Group on tech unions, and I realized that's exactly what my mentor was talking about.

What issues are motivating folks to unionize at The Times?

Honestly, the reasons are things probably common to all workplaces — for example, pay equity is a big one. The News Guild (which represents the newsroom side of The Times) conducted studies that found statistically significant disparities in compensation based on race and gender in the newsroom. At the Tech Guild, we know of disparities anecdotally, but a union would give us more power to address it and force management to take it seriously.

We're also asking for benefits, including parental leave, care for miscarriages, or other fertility concerns; fair promotion processes that are less dictated by your relationship with your manager or your connections within the company; making remote work a long-term possibility to accommodate folks with disabilities; and more robust (and binding) diversity, equity, and inclusion (DEI) policies.

On a more tactical level, how did you all do this? What does a unionization process look like?

We decided that the Tech Guild would encompass everyone who builds, supports, and maintains the technical systems at the Times. In strategic terms, we were trying to carve out a part of the company that makes sense together as a unit, and this is what we came up with. In total, we have around 700 people eligible for the union: the NLRA dictates that no one with hiring or firing power can be in a union, so it's basically all the ICs (to put it in tech-speak).

We have an organizing committee, which makes up around 10% of our bargaining unit (people who are eligible for the union). As an organizing committee member, our job is to reach out to eligible people to hear their workplace concerns, and figure out whether those are things that a union might help with. So that list of motivating issues I provided was the product of hundreds of one-on-one conversations we were all having across the company — not something we arbitrarily came up with.

The goal of these conversations with coworkers, other than soliciting feedback, is to see who might be interested in signing a union card, which acts as an informal vote 'Yes' on the union. Once we pass the threshold of 50%-plus-one 'Yes' votes, we have the legal right to form a union.

In media unions, it's common to ask for voluntary recognition from the company, which is what we tried to do. Unfortunately, NYT management refused. As a result, we have to go into an NLRB election, which is a more formal secret ballot which requires 50%-plus-one 'Yes' votes out of all eligible workers. NLRB elections take time — four to six months — and from the management's perspective, that's more time for the company to inform people about the downsides to the union, or to hope that people will get exhausted thinking about the union for such a long time. In my opinion, this is the company's decision to drag this on longer than it needs to — we know we have the support that we need.

Why are you excited about the union? Why makes it so meaningful?

Where do I begin? For one, I always want to be involved politically in some way, and I've realized that doesn't necessarily have to be something external or policy-focused. Building power within your workplace is a totally legitimate way to enact change that affects not only you and your coworkers, but the industry in general.

For example, last week, we participated in an internal panel with folks from several other media unions about how they were able to develop DEI terms to put in their contract. Time Magazine was one of the first — they mandated in legally binding language that 50% of interview candidates have to be people with identities that are underrepresented in the newsroom. They shared that strategy with us and the other media unions, so we're all now trying to replicate that language in contracts at our own companies.

Something that I sometimes felt insecure about was that as a tech worker, the bread and butter issues aren't so important — I'm not fighting for a living wage, for example. But organizing creates the opportunity to create solidarity across workers at different companies within the same industry, or workers in different roles within an industry. This connection is valuable in and of itself — for people to be able to see their position in the workforce not just as fancy technical people, but as part of something bigger. Having a one-on-one conversation with a fellow worker who might not have thought about any of this before can be really powerful. People fundamentally care about each other, and exposing them to inequities that are impacting people they know and respect and interact with every day is a good way to introduce them to broader, systemic issues.

I think the tech industry is like ripe for unionization, and I think unionizing can be a radicalizing force. Exposure to workers' issues is a necessary counter-action to the Silicon Valley libertarian vision of "every man for himself," of only doing things that are immediately and materially good for me. A lot of problems in tech come out of people not thinking about how things affect others, so thinking more deeply about about solidarity can probably do a lot.

More concretely, unionization is a hub for people who might not typically cross paths to find solidarity with one another, and that's important too.

Subscribe now

Keep up with Angie and the NYT Tech Guild on Twitter.

🌀 microdoses

• 🗞️ The NYT Tech Guild follows a wave of media unions forming over the recent years. This article explains the history and rationale.

• 🌳 Clean your timeline with this profile of Max Falkowitz, who writes a lovely newsletter about learning bonsai from a NYC apartment.

• 🎶 How the internet changed what it means to ~vibe~.

• 🤝 How does a commons differ from the market paradigm we have in the status quo? This table explains. (ICYMI, Reboot's last essay covered commons-based creation.)

• 🔢 Hacktivists rise up

• 📣 Called out

💝 closing note

I asked the Reboot community for recommendations on sci-fi films with a human emphasis (think Her, Arrival, or Eternal Sunshine of the Spotless Mind):

• Anh: I highly recommend Gattaca! It’s so underrated and captures the sci-fi/existential questioning thing well.

• Ben: Have you seen Contact? It inspired Arrival and it’s so good.

• Theresa: I really like this ~indie~ sci-fi short film, World of Tomorrow.

• Lucas: Ex Machina is one of my fav sci-fi-y things — I don’t know if it’s soft-sci-fi, but it's just really well done.

Keep up the good fight,

—Jasmine & Reboot team

I think a lot of technologists have something similar to my Ender's Game moment. They come across a creative work—a book, a video game, a movie—that shocks them into imagining the wild utopias and dystopias that technology makes possible.

Today, I'm sharing an interview with my good friend gubsheep.1 He's a 2020 Thiel Fellow currently building the Dark Forest video game and other zero-knowledge cryptography experiments. Inspired by our last event with Cory Doctorow, I hit him up to chat about his sci-fi inspirations and the design challenges around Dark Forest. Gubsheep also helped create HackLodge, a program for student developers, and he spends his time thinking about crypto, communities, and digital worlds.

🗺 Worldbuilding on the Digital Frontier

This interview has been edited for length and clarity.

What was the first sci-fi book you read that really changed the way you approach your work?

The seeds of my interest in decentralized technology were planted about five years ago, when a friend gave me a copy of True Names: and the Opening of the Cyberspace Frontier. It contains the novella True Names by Vernor Vinge and a collection of essays from cypherpunks and other thought leaders in early cryptography.

Vernor Vinge was one of the first authors to hypothesize about a digitally interconnected world many years before the early internet. True Names lays out a vision for pseudonymity and anonymity in cyberspace; for example, the idea that because of privacy-preserving cryptography, cyberspace may be much harder for traditional meatspace authorities to exert control over. Like other influential sci-fi books like Snow Crash and Ready Player One, the most powerful people in True Names are totally unassuming in the context of traditional power structures—but wield enormous power in a digital reality. Vinge inspired a lot of writers and technologists to start actually building towards this vision.

So tell me about what you're working on now!

For the last 18 months, I've been working on a project called Dark Forest. It's a massively multiplayer, real-time strategy game built on the Ethereum blockchain. It uses new cryptographic technology to make much richer and more dynamic online worlds than were previously possible.

Around 2019, there was an explosion in a subtopic of crypto called zero-knowledge cryptography (ZK crypto), which allows you to perform verifiable computations without showing what data you're performing the computation on.2

Right now, there are two applications of ZK crypto that people care a lot about. One is privacy tech, like sending anonymous cash back and forth on the blockchain. The second is scalability tech: there are ways to use the mathematical techniques behind ZK crypto to make blockchains more efficient. But the third application that I'm most interested in—and that there's been very little activity on so far—is using ZK crypto to create games and digital worlds.

Can you explain me how ZK crypto works in the Dark Forest game?

All games can be divided into two categories: complete information games and incomplete information games. Complete information games are games where everybody knows everything going on, like checkers or chess. Incomplete information games are games like poker or Starcraft where not everybody has full data on the world state. For example, I don't know what's in your hand in poker, and you don't know where my base is located in classical strategy games.

This distinction matters because incomplete information games often have a much richer and deeper strategy space and more interesting social dynamics to explore. If not every player knows the same thing, you enable emergent social dynamics like deception and conditional coordination.

But blockchains by design have open and transparent data layers. On the Bitcoin network, you can see a full record of all the transactions that have happened since the beginning of the system. So until ZK crypto was productized, the only games built on decentralized systems were very simple ones like CryptoKitties and other trading card games.

In 2019, we realized that we could use ZK crypto to make incomplete information games on the blockchain. Now, players can make moves while keeping their private information hidden; that is, a player doesn't need to know exactly where I am to verify that my move was valid.

Dark Forest revolves around a common strategy game mechanic and plot device called the "fog of war." You start off at a home planet in a little corner of the universe. Until you have explored some area of the map, you don't know where or whether other civilizations exist. This anxiety—not knowing what's out there, and not knowing who's seen you—drives a lot of the dramatic action.

How did Liu Cixin's Three-Body trilogy, and specifically The Dark Forest, inspire the storyline for the game?

This answer is fascinating, but contains mild/vague spoilers for the Three-Body trilogy. We've footnoted it, so scroll to the bottom to read!3

Dark Forest engages the central questions of politics, international relations, and human nature itself: for example, do you assume that everybody is power-hungry, or do you think empires can use diplomacy to create cooperative bodies like the UN? Since you don't have a formal background in these subjects, I'm curious how you derive your understanding of these issues.

That's a great question, and my answer is that I don't think we're capable of theorizing about it. We've built a very specific sandbox environment that may or may not bear resemblance to real-world situations. I think the best thing we can do as universe designers is to let people in, see what they do, and try to learn from that.

We've actually been quite surprised by a lot of player behavior. Coming in, we were anchored off Liu Cixin's Three-Body trilogy, where one of the implicit theses is that everyone is maximally suspicious and maximally cautious. Whereas in the Dark Forest game, a lot of players end up engaging in negotiations, trade, diplomacy, and even alliances.

It turns out that the communication channels available to players impacts their behavior. For example, we allow players to link their game accounts to their Twitter handles. One thing that organically emerged is that when players ran into each other, before taking any action, they would DM each other to determine if the other person was hostile, friendly, or might even be open to an alliance to wipe out another bigger player. We later moved the community to Discord and created channels where players can broadcast signals to each other.

The Dark Forest team tries to be fairly hands-off and non-prescriptive about how players should play. I don't think you can stop either players of a video game or players in the grand game of life from doing what they want. We’ve learned that people aren't going to bend to your theorizing.

Do you think that building Dark Forest has made you more optimistic about human nature?

The crypto ecosystem is very polarizing. You truly get the best and the absolute worst of people. For example, Satoshi and Vitalik are so committed to a vision of open technologies, and they've given up so much in the process. But any system that has economic incentives built in will simultaneously attract the most amoral people: the scammers, the grifters, the pump-and-dump shills.

We've been trying to be deliberate about constructing a community that self-selects as the former. So far, there aren't any financial mechanisms built into the game, though in the future, there almost certainly will be. And I've been pleasantly surprised that the game attracts developers who want to write code and share it with others—a lot of our third-party developers spend hours a day helping newbies write scripts or understand the Dark Forest plugin system. So this project has actually made me very optimistic for the decentralized universe community.

You can find more of gubsheep on his Twitter or Dark Forest's Twitter and blog.

Sign up for Reboot

🌀 microdoses

• 🎭 Here's the full text of Vernor Vinge's True Names (though minus the accompanying essays in the book).

• 🌏 Gubsheep also recommended this New Yorker profile of Liu Cixin, which gets into the sci-fi author's upbringing, modern Chinese politics, and his Machiavellian—or pragmatic—imagination. “I did not begin writing for love of literature,” he says. “I did so for love of science.”

• 🕸️ Ethereum creator Vitalik Buterin breaks down the three types of decentralization: architectural, political, and logical.

• 🏰 In this podcast, MacArthur genius and fantasy author NK Jemisin teaches Ezra Klein to world-build. I was especially impressed by the way her historical knowledge informs the constraints of her worlds.

• 🤓 Soulja Boy tell ‘em

• 💰 The following is provided for informational purposes only, and should not be relied on as legal, tax, or investment advice. (Or should it?)

💝 closing note

I've never been a gamer, so I turned to the Reboot Fellows to ask why they game:

• Khoi (mentor, Fanhouse founder): Whereas fiction feels like worldbuilding via relationships and imagination; games feel like worldbuilding via roleplay and mastery. Also, when I used to make games, I'd find people do completely unexpected things. It’s harder to interpret fiction in a novel way from the author's intent than to find delight in a game that was unintended by the game designers.

• Lucas (fellow, Brown 2023): Not my quote, but I thought of Bushnell’s Law, coined by the founder of Atari (Nolan Bushnell): “All the best games are easy to learn and difficult to master. They should reward the first quarter and the hundredth.”

Finally, if you could fill out Reboot's 1-minute feedback survey, that would be awesome. I personally read every single response!

To the moon,

—Jasmine & Reboot team

Disclaimer (Aug 2021): Reboot has since received a $5k grant from Dark Forest’s associated foundation. This newsletter piece was completed long before we asked for or received funding of any kind — in fact, the grantmaking body did not yet exist.

Theresa Gao (she/they/她) spent years as a software engineer at companies like Stitch Fix and Twilio before changing her major to Ethnic Studies and dedicating herself to activism. She was on the founding team of the student #NoTechforICE movement and has worked as a digital organizing fellow at the Chinese Progressive Association and Asian Americans Advancing Justice. She's passionate about work in disability justice and movement building tech.

(We're also testing out an interview instead of an essay today—reply to let me know what you think of this format!)

⚒️ it's time to build (a movement)

This interview has been edited for length and clarity. JS = Jasmine Sun, TG = Theresa Gao.

JS: How did you get into tech? Why was that appealing to young you?

TG: I learned to code in a diversity program, so my technical background was shaped by the fact that I was surrounded by technologists who came from marginalized identities, and who were therefore motivated to make impact.

That changed as I got more into it. I first turned to hackathons because I really liked the idea of building something you're passionate about—there's not that much time to really work on random ideas you're interested in. But over time, I started to realize that hackathons themselves are super corporate spaces.

JS: So what was your radicalizing moment?

TG: I feel like radicalization is less a moment and more of a process. It started with my family—my dad was always really political, he took me to SF Pride when I was a kid—so it was nice to have exposure to political conversations early on.

One of the pivotal points in that process was developing some cynicism about diversity: I started out coding in a diversity program, so I always thought these programs were important. But once I spent time at a tech company, I saw that no matter how diverse the company was, it didn't matter if you don't actually support your workers or create environments that care about them beyond their ability to produce labor.

This made me realize it's not about a company, it's not about statistics—it's about a system.

JS: That's a good place to start talking about NoTechForICE—you were super involved at Stanford. But I feel like the movement often comes down to where a person individually works. How do you frame that campaign in a systemic context rather than blaming individuals?

TG: NoTechForICE pushed me to look past where do I decide to work, which is a temporal, individual problem, to broader questions like why did these companies exist in the first place? What are the forces that are driving the need for this in the market? It’s not a black and white question, there aren’t “good” and “evil” places to work, and that's why shaming individuals isn't the way to go.

During the campaign, I would go to career fairs, talk to people, and hand out pamphlets. I had one conversation with someone from Texas, near the border, and they said I feel really strongly about what's going on with ICE, but I had no idea that Salesforce had a relationship with CBP. Even though people knew ICE and CBP were bad, they didn't understand how they personally were connected to those agencies. So I think it's important to see ourselves as also part of this broader system—we can be agents of violence while also suffering from that violence.

The exciting thing about organizing with people with more radical backgrounds is how they showed me a future I didn't even know existed. That was part of the pamphlet we passed out at career fairs—what freedom from the constraints of these systems looks like, painting a picture of the world we wanted to see. So awareness of the status quo is critical, but so is realizing that the future is not predetermined, that there are different worlds that you can build and be a part of, that you can and should have a stake in that. Or—you make the choice to build a world that destroys those dreams.

We might be students now, but we have agency. I saw Boots Riley at a talk for Sorry To Bother You at Stanford. He said the most powerful thing you can leverage in capitalist society is your labor because your labor is what makes you valuable to a capitalist society. You can choose to work somewhere specific, for example, but the most important thing is understanding where you are located in these systems, where your power comes from.

But at the end of the day, all of this is so much bigger than where any individual chooses to work; this is about imperialism, U.S. history, violence against the Global South. And understanding that is critical.

JS: How does your tech background play a role in your organizing?

TG: The first thing is trying to be conscious and humble about where I come from. Having a tech background doesn't necessarily make me a better organizer; it just gives me a specific perspective.

One thing I noticed is that organizers have powerful messages, but they need to be translated online because not everybody can go to a meeting. Organizing can be really insular, but what's amazing about technology is that it makes movement building inclusive.

I also realized the skills I picked up from hackathons actually translated really well into organizing. For example, activists might be struggling with communication platforms. They might want to build a website or have a media presence. On one level, even knowing that platforms like Signal exist can be super helpful, without writing a single line of code. You know things you don't even realize are important, but first, you need to figure out what people are struggling with.

JS: Let's say you have someone without organizing experience, but who has been reading about this stuff and wants to do something. How would you recommend they start?

TG: There's a lot of hubris in tech. Organizing is about the community, not what you individually can do. You should be asking: how do I, as an individual, support the needs of the community? So educate yourself and prepare yourself, rather than shouting into the void, Hi, I know react.js, I'm here to save the world.

On the other hand, people aren't going to reach out to you. So much is dependent on making that first jump, but you need to understand your positionality. It's not about your skills, it's about the work of creating a better world. There's so much you'll learn from other people that you don't realize. It's a reciprocal relationship.

JS: On the other hand, I imagine a lot of people who have just started learning about activism and justice recognize that other people have been working on it for years. And because these communities often seem insular, a newcomer might be scared of saying the wrong thing or asking dumb questions. How should you approach learning once you're aware of your limits?

TG: It's really hard, and I've definitely been there. Super leftist circles can sometimes get exclusive in a bad way: they intimidate people who would be important to have in the movement, but who get turned off.

That said, there are so many resources that can be super helpful once you realize you have a lot to learn—like theory, or organizing statements, or even mutual aid funds. Beyond that, you can do your own research, if you're still worried about asking "level one" questions. And at this stage, try to prepare as much as possible on your own without being a burden to the organizers you're trying to become involved with.

But you should also realize that people want to help you—there are so many people who started as baby organizers like me, who stumbled into it. I had mentors who believed in me. That's the amazing thing about community organizing: it teaches you why community is so important.

Finally, maybe the first space you find won't be the best space for you, and that's okay! There are so many different communities and causes and campaigns to work on, and eventually you'll find one that works for you. I'm of the hopeful mindset that there will always be a movement space that adheres to your values, that has empathy for people from all skill levels and backgrounds of radicalization.

🌀 microdoses

• 🎒 Want to get started? Read Mijente's guide to organizing your campus against ICE. And if you need a primer on how tech companies fuel deportations, read this explainer first.

• ✊ Tech workers can join their local chapter of the Tech Workers Coalition: "Guided by our vision for an inclusive & equitable tech industry, TWC organizes to build worker power through rank & file self-organization and education."

• 🥚 🐣 🐤

• 💰 Crypto-bros are essential workers, actually.

• 📚 Follow Reboot on Twitter to participate in book giveaways like this :)

💝 closing note

In solidarity with $GME and r/WallStreetBets, I asked the team about their favorite subreddit:

• Jasmine: For some real no-thoughts-head-empty humor, r/disneyvacation, which pokes fun at dumb WikiHow illustrations. (No, I don't know why it's called that.)

• Jihad: Favorite subreddit is r/nba. It scratches my basketball itch, but most importantly, it is the only source of celebrity drama that I allow myself to consume.

• Jessica: As a non-Reddit user, Poshmark is probably the closest thing I use—internet strangers convened for a specific purpose. I selfishly love that it's allowed me to get $15 cashmere sweaters, but more broadly I think the community dynamics are such a fascinating lens on consumption and expectations around the value of goods.

• Ben: Like a lot of people I’ve been doing Zoom DnD since COVID, and DMing my first game. r/DnDBehindTheScreen has been an invaluable and fun resource to have as I learn what I’m doing.

• Jordan: r/delugeusers because of how supportive and kind the community is for helping others learn more about how to integrate the Deluge—a standalone synthesizer, sampler and sequencer—into their music creation workflow.

Thanks for reading!

—Jasmine & Reboot team