The Non-Computational Nature of Agency

The question haunts every discussion of artificial intelligence: can machines have real agency, or do they only simulate it through sophisticated pattern matching? John Vervaeke, cognitive scientist at the University of Toronto, has spent decades developing a framework that answers this question definitively, and his answer illuminates why certain technologies preserve human freedom while others systematically destroy it. His theory of relevance realization reveals agency as non-computational, emerging from processes that resist algorithmic capture because they require something machines lack: care about continued existence.

This insight transforms our understanding of technologies like Bitcoin and Nostr, which succeed precisely because they implement non-algorithmic processes that preserve agency while algorithmic systems only simulate it. Where traditional platforms attempt to compute relevance through engagement algorithms, where central banks attempt to compute value through monetary policy, and where social networks attempt to compute identity through verification systems, Bitcoin and Nostr sidestep computation entirely. They create conditions where agency emerges from thermodynamic work and cryptographic sovereignty, implementing technologically what Vervaeke demonstrates biologically: agency requires processes that transcend algorithmic capture.

Vervaeke's framework begins with a deceptively simple observation: organisms face a combinatorial explosion of potentially relevant information, yet somehow zero in on exactly what matters for their survival and flourishing. A frog ignores thousands of visual details to snap at precisely the flying insect, a human manages complex social situations by attending to subtle facial expressions while filtering out background noise, and even bacteria orient toward nutrients through chemical gradients that require distinguishing signal from noise. This capacity, which Vervaeke terms relevance realization, resists algorithmic formulation because algorithms require pre-defined problem spaces, yet relevance realization generates the problem space within which solutions become possible.

The mechanism operates through what Vervaeke calls opponent processing, where competing cognitive strategies play against each other in dynamic tension. The brain simultaneously explores new possibilities and exploits known patterns, maintains both focused attention and peripheral awareness, seeks both efficiency and resilience. These opposing forces generate an evolving dance, responding to changing contexts in perpetual tension, and this dance emerges from the organism's fundamental care about its own existence. A machine can process data about temperature, but only a living being cares whether it freezes or burns, and this care transforms raw information into relevant information, creating salience fields that guide action through embodied caring.

The implications devastate computational theories of mind that dominate artificial intelligence research. No amount of processing power can solve what Vervaeke identifies as the frame problem: determining which facts matter in any given situation requires already knowing what kind of situation you're in, but recognizing the situation requires knowing which facts to attend to. Algorithms break against this circularity, implementing ever more complex heuristics that fail the moment contexts shift beyond their training distribution. Living beings solve it through embodied interaction with their environment that transcends computation, where caring about outcomes generates relevance without explicit rules.

The frame problem that breaks artificial intelligence reveals why organisms succeed where algorithms fail. Vervaeke demonstrates that relevance emerges from caring engagement with specific environments, a process that resists formalization because it generates the categories it operates within. This explains why decades of AI research pursuing ever more sophisticated algorithms consistently fails to achieve real understanding. The organism's metabolic investment in its own survival generates salience landscapes that no amount of computational power can replicate.

This non-computational nature of agency parallels how value emerges in markets through individual subjective preferences that cannot be centrally calculated. The socialist calculation debate finds new expression in debates about artificial general intelligence, where technologists repeat the same category error: believing that sufficient data and processing power can replace distributed, embodied processes. Just as prices emerge from countless voluntary exchanges without central computation, relevance emerges from the organism's caring interaction with its environment without algorithmic processing.

Bitcoin's proof-of-work consensus mechanism embodies relevance realization at the level of distributed systems, converting energy into computational work that maintains order against entropy. Miners expend electricity searching for valid hashes below a target difficulty, creating blocks that preserve the ledger's integrity, and this thermodynamic process cannot be simulated or shortcut through clever algorithms. The energy expenditure itself constitutes the security model, just as an organism's metabolic investment in maintaining its boundaries against dissolution constitutes its agency. Critics who bemoan Bitcoin's energy usage fail to grasp that this expenditure precisely parallels how living systems maintain coherence: through continuous work against the universal tendency toward disorder.

The protocol achieves what Vervaeke calls autopoietic organization, maintaining itself through its own operation while remaining open to environmental interaction. Each node validates transactions according to rules that emerge from collective agreement yet cannot be imposed by any central authority, creating a system that exhibits real agency at the network level. The blockchain discovers the correct state through thermodynamic work that makes falsification prohibitively expensive, transcending algorithmic processing. This mirrors how organisms don't compute relevance but realize it through metabolic investment in attending to what matters for survival.

Nostr implements a different aspect of agency preservation through cryptographic identity that users fully control. Unlike platforms where identity exists at the pleasure of corporate administrators, Nostr users generate key pairs that constitute their digital existence without requiring permission or validation from any authority. This self-sovereign identity enables what Vervaeke calls participatory knowing, the deepest level of cognition where beings know through transformative engagement. Signing a message with your private key on Nostr goes beyond simple authentication; you participate in creating the social reality of the network through cryptographic proof that cannot be faked or revoked. The protocol preserves agency by preventing the algorithmic capture that occurs when platforms intermediate human relationships through engagement metrics and recommendation systems.

The convergence of Vervaeke's cognitive science with Austrian economics and cypherpunk technology reveals a profound truth about human freedom in the digital age. Agency resists computation, optimization, and algorithmic enhancement; it can be preserved or destroyed. The technologies that matter for human flourishing are those that protect the non-computational processes from which agency emerges, transcending speed and accuracy metrics. Bitcoin preserves economic agency by making monetary control thermodynamically expensive, Nostr preserves communicative agency by making identity cryptographically sovereign, and both succeed precisely because they don't attempt to compute what cannot be calculated.

Understanding agency as non-computational explains why certain technologies feel liberating while others feel oppressive, why decentralized systems enable human flourishing while centralized platforms create dependency and despair. The meaning crisis that Vervaeke diagnoses and the freedom that cypherpunks defend share the same root: humans need agency to create meaning, and agency requires escaping algorithmic capture. Platforms that reduce human interaction to engagement metrics, governments that reduce economic activity to GDP statistics, algorithms that reduce relevance to click-through rates: all erode the foundations of meaning-making that make life worth living.

The path forward requires building technologies that respect the non-computational nature of agency, that create spaces for relevance realization, value emergence, and identity expression without algorithmic mediation. Bitcoin and Nostr point toward this future as existence proofs that such technologies can exist and thrive, pointing toward deeper possibilities. They demonstrate that we can build systems that preserve agency without requiring trust, that enable coordination without central control, that facilitate human flourishing without algorithmic intermediation. The question is not whether machines can have agency, but whether we'll build technologies that preserve the agency we already possess.