What’s in an Ecosystem?
Taking the ecosystem metaphor seriously in the digital world
Two quick Kernel-related notes:
First, we’ve unlocked our first post-launch essay, “Democracy-Affirming AI Could Make Things Even Worse” written by Kevin Klyman and edited by Elisabeth Siegel. It’s very carefully researched — hope you’ll give it a read!
Second, if you’re in SF, we’ll be at Lit Crawl next Saturday 10/21! We’ll be reading at Noisebridge from 6:30-7:30pm. Hope to see you there!
Ecological metaphors are everywhere in the digital world, but how can we be more intentional and learn from them? Today, let’s dive into ecosystems!
What’s in an Ecosystem?
“The outstanding scientific discovery of the twentieth century is not television, or radio, but rather the complexity of the land organism. Only those who know the most about it can appreciate how little we know about it.” - Aldo Leopold, Round River (from 1948, pub. 1993)
Our characterization of the digital world is replete with ecological metaphors, and more specifically, the metaphor of the ecosystem. The term “ecosystem” has expanded far beyond its original reference to organisms and their interactions within their natural environment and now refers to topics as diverse as data, open source software, and Apple’s App Store. The ecosystem metaphor is not only routinely invoked by venture capital firms and management consultants, but utilized by White House memos, government funding programs, and bills in Congress. Others have taken the metaphor to its logical end, analogizing AI harms to “oil spills in the information ecosystem” and atmospheric pollution in the “AI ecosystem” that leads to “internet climate change.”
The ecosystem metaphor is easy to invoke, yet bridges two concepts that couldn’t be more different. While natural ecosystems are made of living organisms in their environment, digital ecosystems are constructed spaces designed to hold bits. Yet using the metaphor provides a certain allure, and I’ve often used it myself. It seems natural, and indeed almost automatic, to invoke our conceptions of complex natural systems to better reckon with complex information systems—especially those with a variety of actors and stakeholders with interweaving relationships, where a single action may have cascading consequences.
What does a more intentional approach to the ecosystem metaphor look like? An ecosystem is more than just an indicator of complexity. The digital world can learn much more from the legacy of ecologists, conservationists, and traditional and indigenous cultures in interacting with natural ecosystems. At the same time, embracing the artificiality of the digital world—the very fact that sets it apart from nature—provides opportunities to more ambitiously and effectively govern it.
The Origins and Rise of the Ecosystem
The very origins of the term “ecosystem” demonstrate its limitations. It begins in 1866, when German zoologist Ernst Haeckel coined the term “ecology” to refer to the “economy of nature”: the relations between an organism and its “inorganic and…organic environment.” That “ecology” comes from “economy” hints that the field is itself an abstraction, a constructed metaphor over more amorphous and complex forces of nature. Seven decades later, Arthur Tansley coined the term “ecosystem” to emphasize that the study of the environment, at its core, involves human-created abstractions: systems. Ecosystems, in Tansley’s words, “range from the universe as a whole down to the atom” and “overlap, interlock and interact with one another,” as they are artificially and somewhat arbitrarily defined. The boundaries of an ecosystem delineate living communities, their environment, and the relationships between both. Like any abstraction, the ecosystem concept is incomplete and has its limits: it only captures a subsection of the natural world.
The post-WWII era saw a move towards making sciences more quantitative and individualistic. This greatly influenced ecology, with fields such as population dynamics and statistical ecology increasing in popularity. Information theory and ecology became intertwined: Shannon’s information entropy index is also used for biodiversity. This era saw such ecologists as Eugene Odum, who studied quantitative ecosystem dynamics; popularized the concept of ecosystems through a seminal textbook; and helped make ecology a discipline in its own right. Later, pioneers such as Gregory Bateson would connect ecology, computation, and complex systems through cybernetics. In this era, ecology was also shaped by the rise of Big Science: massive government-funded investment around science and R&D, from cyclotrons to atomic weapons to pesticides. Some of Odum’s most important work was funded by the Atomic Energy Commission to study the environmental impact of nuclear weapons production.
In the ‘60s, the excesses of scientific advancement, and a reckoning with its dangerous externalities, led to a resurgence of the American environmental movement that focused on a more holistic view of systems, with such prominent figures as Rachel Carson and Ralph Nader. But the environmentalist movement also inspired a generation of technologists who would go on to transform our digital world. Stewart Brand’s Whole Earth Catalog not only brought together environmentalists and technologists, but helped inspire the Homebrew Computer Club, and, ultimately, the roots of modern computing and the Internet as we know it. And ecological science fiction such as Dune reimagined the relationship between ecosystems and technology. This relationship often took the form of resistance. In the ‘70s, three women health and labor rights organizers—Robin Baker, Amanda Hawes, and Pat Lamborn—formed the Santa Clara Center for Occupational Safety and Health (SCCOSH) to advocate for health and safety for Silicon Valley’s working-class minorities. In the ‘80s, the Silicon Valley Toxics Coalition (SVTC) began dealing with such issues as groundwater contamination by high-tech manufacturing facilities and e-waste.
Ecological terms also began expanding to different fields of study. In the ‘60s, anthropologists began measuring calorie exchanges in human societies through “human” or “cultural” ecology. In 1968, Neil Postman coined “media ecology” to build on Marshall McLuhan’s ideas of how media affects the mental environment, and the ‘70s saw the development of “information ecology” to recognize the interdependence among parts of an information system. These uses of “ecology” seem to have paved the way for “ecosystem” to surpass it as the dominant metaphor in the digital world in the twenty-first century. Though all these approaches were certainly inspired by ecological ideals, did they really take the metaphor—and our ecological legacy—seriously?
Taking Ecosystems Seriously, Part 1: Learning from Nature
“Now, let us consider for a moment the question of whether a computer thinks. I would state that it does not. What ‘thinks’ and engages in ‘trial and error’ is the man plus the computer plus the environment. And the lines between man, computer, and environment are purely artificial, fictitious lines … What thinks is the total system” - Gregory Bateson, Steps to an Ecology of Mind (1972)
In an ecosystem, actors are not separate from their context. The classic example of a food web shows how natural ecosystems comprise not just the organisms contained in them, but how they connect to each other and their environment. Such a viewpoint opens up new areas for inquiry in the digital space. For example, when asking if an AI model should be widely available—or if self-driving cars should be permitted—we must also ask: who is building the model, and what are their incentives? Who are the intended users, and who might be subject to collateral damage? What legal structures affect how it is deployed and who pays for the harms? A detached analysis of a new technology, without its context, is simply incomplete.
As with the natural world, approaching the digital world is also ultimately a choice of stewardship versus exploitation. We can draw upon traditions of indigenous stewardship of the environment to provide more healthy and sustainable approaches to the digital world. For example, the Abundant Intelligences project reimagines how to design AI using indigenous knowledge systems, grounded in methods that focus on “abundance rather than scarcity” and “recognize the abundant multiplicity of ways of being intelligent in the world.” Or consider religious attitudes towards nature, for example the Hindu prayer of asking forgiveness to Mother Earth even for touching her with one’s feet. When the natural environment is treated with such respect and conscientiousness, it is clear that components of an ecosystem are not merely instrumental. In the digital world, we should “think of objects not as instruments for our use, but as entities that are effectively linked and that need care - to think of objects as plants in our garden.”
We should thus be more careful about polluting the information environment with bad data, and be more purposeful about investing in and sustaining, rather than freely using or exploiting, public goods such as free and open source software. Keep in mind that the end state of natural systems is not to stay constant, but rather a state of dynamic equilibrium, where systems are resilient and can deal with change. When we build digital systems, we should aim not for efficiency, but rather sustainability. In fact, more efficient systems can end up being more brittle, and conversely, adding friction can make systems more safe and resilient.
Finally, natural ecosystems teach us to be aware of their limitations. The natural environment has taught us the dangers of ignoring real harms caused by exploitation of a seemingly limitless environment, from fossil fuel extraction to overfishing. In the digital world, these harms include the environmental costs of compute, labor costs of automation, and the harms caused by misinformation. The digital world is not a frontier or a wilderness to be colonized or disrupted, but each step taken in the digital world should be mindful of its impacts on society.
Taking Ecosystems Seriously, Part 2: Embracing the Artificial
“He sensed planning for centuries ahead: trees to replace building beams, to hold watersheds, plants to keep lake and river banks from crumbling, to hold topsoil safe from rain and wind, to maintain seashores and even in the waters to make places for fish to breed.” - Frank Herbert, Chapterhouse: Dune (1985)
The Internet is a constructed space, made of negotiated standards and allocated domain names. Unlike the natural environment, everything in the digital world is ultimately code that was designed, and written, by somebody. This is where the ecosystem metaphor breaks down, and we must take care to recognize this difference. Digital spaces are not to be passively observed, but constantly co-created. And by embracing the artificial and constructed nature of digital spaces, we empower ourselves to participate in their co-creation.
We can design and shape our digital environment to be less toxic, more collaborative, and more community-driven. As Zizi Papacharissi notes: “Toxicity is a human attribute, not an element inherent to digital life. Unless we design spaces to explicitly prohibit/penalize and curate against toxicity, we will not see an improvement.” For example, this may look like prioritizing “good” over growth; being more intentional about what content we add to the digital commons; or building in values into the design process of new technologies. In the digital world, we not only have the ability, but the imperative, to design and help construct such spaces. Technological developments are neither neutral nor inevitable, and our tools for shaping them are manifold: legal frameworks, political movements, lines of code, web standards, and design patterns.
The artificiality of the digital world also means it should ultimately be subordinated to the natural world, which is irreplaceable. We should not fall into the trap of using the ecosystem metaphor to trivialize or minimize real-world impacts. Terms such as “cloud computing” often give a nice natural analogy while obscuring the reality behind them: data centers that eat up huge amounts of energy and resources from the environment. The frictionlessness of crypto belies the environmental impact of mining. And the ease of accessing AI models obscures the amount of electricity—and water—needed to train and run them. These externalities remind us that technology is only a means to an end, and that end is to serve people. Technological progress should only be secondary to, and ultimately be accountable to, broader societal goals and democratic values.
Conclusion
It’s worth noting that perhaps the difference between the natural and digital worlds isn’t as pronounced as it seems. The digital world, though it deals in virtual abstractions, does so on behalf of living beings—after all, the ultimate users of the Internet are people. And the natural environment is not untouched by humans, but constructed to some degree—even setting aside anthropogenic climate change, humans have been shaping the environment for millennia. But these overlaps only emphasize the importance of being more careful and intentional about the use of the ecosystem metaphor.
Metaphors aren’t just carelessly tossed around—they shape ideologies. An automatic acceptance of the ecosystem metaphor can lead, at the worst, to an unbounded attitude of extraction and harm in our digital world; a resigned approach to technological inevitability; and disregard of its consequences. By taking ecosystems seriously in the digital world, we demand better. First, we draw from traditions of stewardship and respect of the natural world. Second, we embrace our ability to shape a world that is continually constructed, designed, and shaped by our actions. In doing so, we will not only co-create a better digital world, but help preserve and sustain a healthy and resilient natural environment for all.
Ashwin Ramaswami is an editor of Reboot and works on law, policy, and strengthening the open source software ecosystem.
Reboot publishes essays on tech, humanity, and power every week. If you want to keep up with the community, subscribe below ⚡️
🌀 microdoses
Externalities? What externalities? “From an ecological perspective, there is no such thing as an externality, because there is no outside, to which effects (such as pollution or social inequality) could be externalized” - from Attila Márton
Startups and founders use “ecosystem” a lot too! The myth of the heroic founder belies a combination of institutional forces and teams that make big things possible.
💝 closing note
Thanks for reading! Go steward the digital ecosystem ❤️
Ashwin & Reboot team
Thanks, great article! I like how you phrased "technology is only a means to an end, and that end is to serve people". However, it seems we have entered a period where technology has such a power that, whilst it needs to fade into the background, we have entertained it to a place where we have instead put it on a pedestal. As such we often make technology the end by mistake.