Most people working at the frontier of anything genuinely important right now share a specific, uncomfortable experience: the sense that the work that matters most — the work that is irreducibly contextual, relational, and hard to explain — is invisible to institutional incentive structures. Researchers who follow genuinely novel threads get penalized by metrics optimized for legible consensus. Builders who prioritize trust and care over throughput lose funding rounds to faster-moving competitors. Artists, educators, and community organizers watch their contributions get classified as "externalities" by the very systems their work keeps alive. Meanwhile, the recommendation algorithms that were supposed to connect us to what we need have collapsed the information landscape into a low-dimensional tube that preferentially surfaces outrage and anxiety, gradually narrowing the adjacent possible available to each of us and eroding the cognitive diversity that any healthy collective intelligence depends on.

The frame this talk offers is a structural account of why these experiences are happening simultaneously, why they are symptoms of the same underlying dynamic, and how to see the very existence of the DWeb ecosystem as evidence that we are on track for a total inversion of economic value.

Drawing on twenty years of research into complex adaptive systems and major evolutionary transitions, Michael Garfield (formerly Santa Fe Institute, Long Now Foundation, Mozilla) will recast our current economic, ecological, and epistemic crises as a transitional moment in the emergence of The Living Web: a future of thriving collective intelligence enabled by decentralized sensemaking infrastructure.

The key is a fact worth sitting with: only 1% of animal DNA codes proteins. The other 99%, dismissed for most of the 20th century as "junk", handles epigenetic regulation and serves as a standing reserve of adaptive potential that makes complex life possible. Research on the transition between bacterial and nucleated cells locates this as a discontinuous jump at which noncoding sequences suddenly appear and the trajectory of life on Earth changes permanently. And the metabolic data tells the same story: as bacteria grow larger, an increasing fraction of their energy goes to growth and production. At the transition to complex cellular life, that relationship inverts. Larger organisms devote more of their metabolism to maintenance, regulation, and the sensemaking layer. This inversion is physics, a necessary consequence of the information processing demands of a more complex system.

The industrial economy has been a prokaryotic economy — nearly all coding sequences, optimized for throughput, treating the work that resists quantification as overhead. The norm of "product-market fit" says: your value must be legible to the market's existing vocabulary, or it doesn't count. This is a coherent strategy for a simple organism. It is an unstable one for a system whose information processing demands have outrun what a purely production-maximizing logic can handle. What we experience as the "metacrisis" — the strange convergence of civilizational risks into a single knot — is what model collapse looks like at the scale of an entire economy: a loss of variance so severe that the system can no longer adapt to the novelty it generates. But as near-zero marginal cost automation handles an increasing share of legible, replicable cognitive work, the greatest economic reward begins moving toward whatever cannot be replicated: context-specific insight, divergent exploration, and the maintenance of cognitive diversity in systems that would otherwise converge toward brittleness.

We evolved for richly structured environments that reward open-ended, high-dimensional search — information ecologies in which genuine exploration reliably yields both nourishment and unexpected discovery. The recommendation algorithm is the opposite: a low-dimensional tube that feeds us what the system predicts we already want, gradually narrowing the adjacent possible and eroding, in aggregate, the variance the collective needs to stay alive. But this is a transitional state, creating new selection pressures for the infrastructure we need to operate at the actual scale of 21st century complexity. And under these pressures, we are building toward a future of high-dimensional collective foraging — where sovereign knowledge infrastructure, personal context engines, local-first federated architecture, consent-based selective legibility create the conditions under which “scenius at scale” becomes achievable.

Theoretical work on the evolution of syntactic language gives us the same argument from a different angle. The sentence emerged from a scaling-induced crisis in the internal coordination of prehistoric human societies, one that resulted in an explosion of available vocabulary. Most words only matter in very specific contexts; the sentence is what makes context-specificity meaningful rather than merely idiosyncratic. The industrial economy has been communicating in words: legible units of value designed to mean the same thing regardless of who is reading them. The emerging economy communicates in sentences: value constituted by its context, meaningful inside a specific epistemic community, a specific position in an evolving information ecology. The measure of contribution shifts from "how many people can use this?" to "how specifically does this serve the collective intelligence of the systems I'm embedded in?"

This reframes what the individual is for. The ideal of self-actualization — finding and optimizing a fixed essential self, stable expertise, and legible brand — made structural sense when legibility to a generic market was the measure of value. What the emerging economy selects for not fixed specialists but context-specific explorers, organisms that maintain their adaptive capacity because they haven't fully offloaded it onto the supply chain. And the most valuable thing in this new economy will be divergence into “the adjacent possible” — providing unique signals to collective wayfinding systems that need exactly the variance you embody, given where you are and what you have seen. This is the shift from self-actualization to self-transformation, from mastering a fixed identity to a continuous process of becoming something the system hasn't encountered before. It’s a future where “centering diverse perspectives" is not a value to honor but a functional requirement. The DWeb Principles, read in this light, are not a normative vision of what the Web should be…they are a description of what any sufficiently complex system must eventually become to survive its own scale.

It’ll be the end of “bullshit jobs” and marginalized creative labor…and the beginning of a tech-enabled flock of human and ecological intelligences exploring new horizons for meaningful work.
Intended as a provocation for a week’s worth of discussion, this talk will build the argument from biological first principles to its practical implications, ending in genuine open questions:
• Where are we in this transition?
• What does the current landscape of affordances make possible that wasn't possible three years ago?
• What information architectures and user interfaces will allow people to chart their divergence from low-value legible paths through possibility-space and identify the most salient open questions?

Format: No special equipment needed. Circle seating welcome; outdoors fine.

Speaker: Michael Garfield is a writer and artist focused on transdisciplinary research into the evolutionary history and future of intelligence. He is a co-founder at Atlas Research Group — which builds high-dimensional sensemaking tools on sovereign, commons-oriented infrastructure — and the award-winning host of Humans on the Loop. He has worked with the Santa Fe Institute, Long Now Foundation, Mozilla, and ASU’s Center for Science and the Imagination. He lives in Santa Fe, New Mexico.