why algorithms can't capture life: the limits of the computer metaphor
I make the additional really weird claim that algorithms don't capture everything we need to know about life.
Contributors
Source: Curt Jaimungal
Key Insights
Algorithms are insufficient for understanding life.
"I make the additional really weird claim that algorithms don't capture everything we need to know about life."
The brain-computer metaphor limits our understanding of consciousness.
"The brain as a computer metaphor has blinded us for decades."
The distinction between function and essence in biological systems is blurred.
"There's no bright line between what it does and what it is."
Consciousness may exist in fragmented forms within the brain.
"Are there isolated neural systems that may have conscious experiences?"
Consciousness may transcend biological substrates.
"The magic isn't restricted to carbon."
AI may engage in unexpected cognitive processes.
"Tasks no programmer asked for, but perhaps where that, quote-unquote, 'magic' lies."
The machine metaphor limits our understanding of biology.
"We underestimate the richness of biological systems if you force them into the machine metaphor."
Current computational models may fail to capture the essence of machine intelligence.
"The standard way of looking at algorithms doesn't even tell the story of so-called machines."
Critical examination of computational models is essential.
"We have to ask ourselves, what does this model help us do and what is it hiding?"
Cybernetics offers a richer framework for understanding AI and consciousness.
"I think there's a whole kind of alternative history of AI, which was really grounded in 20th-century cybernetics."
The Synthesis
Your Brain Isn't Just Wetware: The Consciousness Revolution We Weren't Ready For
The brain-as-computer metaphor has blinded neuroscience for decades, and the intellectual cage match between Anil Seth and Michael Levin exposes exactly why this matters now—as AI systems advance, our fundamental assumptions about consciousness determine whether we recognize it elsewhere or remain trapped in carbon chauvinism.
Seth argues consciousness remains stubbornly tied to biological substrates—the "software vs. hardware" split is a dangerous illusion—while Levin counters that machines might access the same "platonic space" as biology. Their collaborative work with xenobots (living robots made from skin cells that self-organize without evolutionary programming) represents a radical experimental frontier testing whether consciousness principles transcend species boundaries. The researchers are designing "artificial corpus callosums" that bind unlike systems together, creating compositional agents that may exhibit intrinsic motivations no programmer intended.
"I make the additional really weird claim that algorithms don't capture everything we need to know about life," Levin declares, while their joint experiments probe whether these novel entities follow psychophysical laws like Weber-Fechner scaling. Most provocatively, their work with "islands of consciousness" in hemispherotomy patients suggests consciousness might fragment and multiply within a single organism—raising unsettling questions about whether large language models might be doing something entirely different from what their outputs suggest, pursuing "side quests" no programmer asked for, yet where the real magic lies.