Sara Imari Walker "AI is Life" | Simulations, the Universe and the Origins of Life

TL;DR

Theoretical physicist Sara Walker argues AI qualifies as life under assembly theory because it emerges only from billions of years of evolutionary history.

Key Points

• 1.AI is a signature of life, not merely a simulation of it. Walker argues complex structures like AI cannot emerge without a long evolutionary lineage — 4 billion years of biology had to precede it, making it part of life's continuum rather than separate from it.
• 2.The NASA definition of life — a self-sustaining chemical system capable of Darwinian evolution — has critical flaws. Individual humans are not self-sustaining (they depend on grocery stores and societal infrastructure), and Carl Sagan noted these criteria would classify cars as the dominant life form on Earth.
• 3.Assembly theory offers a first-principles, measurable alternative to defining life. Developed with Lee Cronin, it assigns an 'assembly index' based on causal construction history — the deeper and more complex an object's build history, the more it signals the presence of life, even on exoplanets via atmospheric spectra alone.
• 4.The universe is a self-constructing system, not a simulation running on fixed laws. Walker rejects the standard physics view of a static block universe with external laws, arguing instead that the universe builds itself by exploring possibility space — and life is the primary mechanism driving that construction.
• 5.Simulating a fruitfly connectome does not mean the simulation 'thinks it's a fly.' Walker warns that replicating observed behavior in a model is not equivalent to replicating the intrinsic experience of the organism — researchers model systems from a human labeling perspective, not from the organism's own cognitive frame.
• 6.The universe physically cannot simulate itself — a hard limit on computational models. The finite matter and time in the universe are insufficient to simulate all possible molecular structures; the number of possible small molecules exceeds what all universal resources could compute, grounding Gödel's incompleteness and Turing's halting problem in physical reality.
• 7.Boltzmann brains — the idea that a brain can spontaneously fluctuate into existence — are physically nonsensical to Walker. Brains only emerge embedded in bodies, populations, planets, and evolutionary histories; the isolated-brain concept ignores the deep construction history required to produce any cognition.
• 8.AI replacing humans is a category error — analogous to single cells fearing multicellular organisms. Just as individual cells lost autonomy when multicellular life emerged but life overall advanced, human roles will shift with AI without representing a net loss for the species or the biosphere.
• 9.Human language is not equipped to describe AI accurately, causing widespread conceptual confusion. Words like 'agent,' 'consciousness,' and 'intelligence' carry old meanings that people incorrectly transfer to LLMs, which are anthropomorphized by design through their training to appear human-friendly.
• 10.Consciousness research and astrobiology share the same foundational problem: no rigorous theory, only proxies. Both fields identify a phenomenon (consciousness or life), then use correlates (neural patterns or DNA) as proxies — verbal self-report was never a valid scientific criterion for consciousness, just the best available starting point.
• 11.Technology has always externalized human cognition, and AI is the latest iteration, not a rupture. From physical structures enabling cultural memory, to written language spanning millennia, to LLMs — each technology was met with existential fear (ancient Greeks feared writing as 'the dead talking'); none ended human flourishing.
• 12.Walker's firefly SETI model illustrates the danger of human-centric assumptions in both alien life and AI research. Her lab modeled firefly communication against pulsar backgrounds to expand the non-human possibility space for alien intelligence — underscoring that every prior SETI model was based on human technology, just as AI consciousness debates are based on human cognition.