{"slug": "the-nature-of-free-will-in-the-age-of-ai", "title": "The Nature Of Free Will In The Age Of AI", "summary": "In an essay originally published in Chinese by San Lian Lifeweek, science reporter Albert Yuan examines the philosophical question of free will in the age of AI, arguing that the feeling of losing free will is a philosophical illusion. He traces three historical crises—determinism, monotheism, and modern science—and discusses the Libet experiment, which showed a readiness potential beginning 550 milliseconds before movement, with conscious awareness arriving 200 milliseconds before movement, suggesting the brain decides 350 milliseconds before the person is aware.", "body_md": "Albert Yuan is a science reporter at San Lian Lifeweek, one of China’s leading news and lifestyle magazines, where this essay was originally published in Chinese.\n\nLet me begin with breakfast. You wake up, shuffle to the kitchen. Coffee or juice? A small decision, trivial. And yet nested inside that trivial choice is one of the most consequential questions in the history of human thought: Did *you* actually make the choice? Or were you — to borrow a phrase I find both evocative and ultimately misleading — simply a puppet whose strings were pulled long before you reached the kitchen?\n\nAs the UCLA professor Pamela Hieronymi told me, “I ask my students this question, and I watch something happen in the room. Not quite panic, but something adjacent to it. There’s a slow deflation, as if I pulled a plug from the bottom of a pool and all the water — all the meaning, all the moral weight, the sense that their lives are genuinely *theirs* — drains away. I have seen this happen while teaching nothing more dramatic than Isaac Newton’s three laws of motion, a discussion about billiard balls and planetary orbits. And somehow the students walk out feeling that their choices no longer belong to them.”\n\nI want to take this feeling seriously. And then I want to explain why I think it is, in an important sense, a philosophical illusion — though not the kind of illusion its critics suppose.\n\n**The Three Crises**\n\nThe question of free will has survived three major crises in Western intellectual history, each one threatening to extinguish it entirely.\n\nThe first was in ancient Greece, roughly 2,500 years ago, when philosophers like Democritus and Leucippus looked at the world with admirable rigor and concluded that everything — atoms bouncing in a void, future events — was already determined. The first crack: determinism. The second crack: the causal chain. Your choice of breakfast traces back not to you but to your caffeine dependency, which traces to your addiction-prone neurobiology, which traces to your parents, which traces to the emergence of mammals, which traces to the birth of the solar system. At some point, the causal thread escapes your body entirely.\n\nThe second crisis emerged with monotheism. If God is omniscient and omnipotent, in what sense are human choices genuinely free? St. Augustine answered with characteristic ingenuity: God gave us the freedom to choose good, but our ancestors squandered it in the Garden of Eden. Thomas Aquinas deepened the synthesis, positioning God as the “first cause” and human rational agency as a “second cause” — not in conflict, but nested. For centuries, this reconciliation held.\n\nThe third crisis is the one we are still living through, and it is the most democratized of the three. It began with Newton, accelerated through Darwin and has now arrived, via neuroscience, in everyone’s living room. The physicist showed us a clockwork universe. The biologist showed us that organisms are the undesigned products of genetic mutation and natural selection. And neuroscience — particularly one elegantly simple experiment — showed us something that feels, to many people, like the final blow.\n\n**The Libet Experiment And Its Discontents**\n\nIn the early 1980s, a quietly obsessive neuroscientist named Benjamin Libet, working at UC San Francisco, designed an experiment so simple it seems almost quaint. He asked volunteers to flex their wrists whenever they felt like it, and to note the position of a rotating dot on a clock face at the moment they first felt the urge to move. Meanwhile, he measured the “readiness potential” — a slow build-up of electrical activity in the brain that precedes voluntary movement.\n\nThe result was disquieting. The readiness potential began approximately 550 milliseconds before any actual movement. The subjects’ conscious awareness of their intention arrived only about 200 milliseconds before the movement. Which means the brain had “decided” to act roughly 350 milliseconds before the person was consciously aware of any intention.\n\nThe popular press seized on this with the enthusiasm that popular presses have for anything that confirms our darkest suspicions about ourselves. *Free will is an illusion*, the headlines announced. *Your brain decides before you do.* And this conclusion rippled outward, amplified by books like Sam Harris’s “Free Will” (2012) and Robert Sapolsky’s “Determined” (2023), which used it as one stone in a much larger avalanche.\n\nSapolsky’s book is, I should say, a remarkable intellectual achievement. It assembles an extraordinary range of evidence — from the effects of testosterone on aggression to the long-term neurological consequences of childhood trauma to the ways in which gut microbiomes influence mood and decision-making — to argue that every human behavior can, if you trace the causal chain back far enough, be attributed to factors entirely outside a person’s control. He writes with the conviction of someone who decided, at age 14, awakened from a nightmare, that neither God nor free will nor purpose exists in the universe.\n\nThe conclusion he draws is practical: We should stop punishing criminals and focus instead on prevention and containment. He points to Norway’s treatment of Anders Breivik — the man who murdered 77 people in 2011 and is serving a sentence of 21 years (with the possibility of indefinite extension) in what amounts to a comfortable apartment with access to university courses and a computer — as something approaching the right idea.\n\nI find myself in partial sympathy with this view and in deep disagreement with its philosophical foundations.\n\n**What Free Will Actually Means**\n\nHere is where I want to slow down, because I think the debate over free will has long been running on a set of confused definitions, and clarity here matters enormously.\n\nWhen people ask whether we have free will, they are usually asking one of two very different questions. The first: *Can we, through our own efforts, live lives for which we are genuinely responsible — lives that have moral meaning?* The second: *Can we, without any prior cause, intervene in the physical world?*\n\nThese are not the same question. The first is the one I care about. The second is the one Newton destabilized, and rightly so. But here is what I think Sapolsky and Harris and others miss: answering “no” to the second question does not settle the first. The questions are on different levels. They are, as Immanuel Kant would have said — though I hold this view without the elaborate metaphysical scaffolding he required — about different dimensions of the same reality.\n\nCompatibilism — the view that free will and determinism are not in conflict — is not, as Harris famously quipped, merely a puppet who has come to love his strings. That characterization misunderstands what compatibilists are actually claiming. We are not denying that causal chains exist. We are asking: Which causal chains matter for questions of agency and responsibility? And our answer is: The ones that run through the person, through their reasoning, their values, their reflective endorsement or rejection of their own desires.\n\nConsider Harry Frankfurt’s insight about the hierarchy of desires. I want a glass of wine. That is a first-order desire. I also want to avoid excess because I value my health and my sobriety. That is a second-order desire. When my first-order desire and my second-order endorsement align — when I drink the wine because I genuinely choose to, having reflected on what I value — that is freedom of will. When they conflict — when an addict craves a substance against their reflective wishes — that is a paradigm case of unfreedom. The question of whether a causal chain can be traced backward to the Big Bang is simply not relevant for this distinction.\n\n**The Neuroscience Is More Complicated**\n\nWhen I first read Libet’s paper in the early 1990s, I thought his interpretation of his own data was wrong. I still think so, and the intervening decades of neuroscience have given me good reason for that view.\n\nThe most important corrective came from Aaron Schurger, a neuroscientist who noticed that the “readiness potential” Libet had taken as evidence of unconscious decision-making looked remarkably like the random fluctuations that occur in any large neural network. When a person is asked to move “whenever they feel like it,” they are essentially waiting for the brain’s background noise to cross a threshold. The readiness potential may simply be the signature of that threshold-crossing event — not the mark of a decision made before consciousness arrived.\n\nUri Maoz at Chapman University pushed further. When he gave subjects a choice that actually *mattered* — which of two charities would receive a real $1,000 donation — the readiness potential disappeared entirely. The brain signatures of meaningful, deliberate choices look nothing like the signatures of arbitrary wrist-flexing. This is not a trivial thing. It suggests that the Libet experiment tells us something interesting about random neural fluctuations but almost nothing about the kind of choice that philosophers and ordinary people actually care about when they ask whether we are free.\n\nThe brain’s left hemisphere, as Michael Gazzaniga’s beautiful split-brain research shows, is indeed a compulsive storyteller — an “interpreter module” that weaves post-hoc narratives about our reasons. We sometimes act for reasons we aren’t conscious of, and then confabulate explanations later. This is genuine and fascinating and important. But it is not evidence that deliberate, reflective, values-based choice is illusory. It is evidence that much of our behavior is not that kind of choice — which is precisely the point that compatibilists have been making for centuries.\n\n**Why The Stakes Are So High**\n\nLet me be honest about something. Part of what motivates the persistent defense of free will — and I think this is legitimate, not merely self-interested — is that the practical consequences of abandoning it are not trivial.\n\nPsychologist Kathleen Vohs conducted experiments showing that subjects who read passages arguing for determinism subsequently behaved more dishonestly and less prosocially. They cheated more on tests. They stole more money from envelopes. Now, one might argue — as Sapolsky does — that this just shows how deeply the free will illusion is entrenched, and that properly educated determinists would not behave this way. Harris himself says that abandoning the belief in free will has made him more compassionate, not less.\n\nMaybe. But I think the relationship between belief and behavior here is genuinely complex, and that there is a real risk in popularizing a view that says: No one is really responsible for anything. Not because the philosophical argument is weak, but because the cognitive and social infrastructure that sustains our practices of holding one another to standards — the institutions of accountability, the language of praise and blame, the interpersonal fabric of trust — may depend on certain commitments that cannot survive the radical eliminativist view.\n\nThis is not, I want to be clear, an argument for believing falsehoods. It is an argument for taking seriously the question of what free will actually *is* before we announce that we have refuted it. If what we have refuted is the libertarian notion of causally unconditioned agency — the idea that you could have done otherwise even when the past is precisely the same — then perhaps we have not touched free will in any sense that matters. If what we have established is that your choices emerge from your brain, your history, your character, your values — well, yes. Where else would they come from? That was never in dispute.\n\n**The Biological Picture**\n\nThe neuroscientist Kevin Mitchell offers a perspective that I find genuinely illuminating. He distinguishes between physical determinism, causal determinism and biological determinism — only the first two, he argues, are in tension with free will. The third — the kind of determinism native to living systems — is precisely what *constitutes* genuine agency.\n\nLiving organisms, he argues, act for their own reasons. They are driven not by energy but by information — information whose meaning is internal to their structure and history. A bacterium navigating toward glucose is not a billiard ball responding to a force. It is integrating signals about its external environment with information about its internal state — how hungry it is, how toxic its surroundings, what its ancestors learned about similar situations — and producing behavior accordingly. This is the embryo of agency.\n\nAs nervous systems grew more complex, this capacity for information integration and meaning-making expanded. The emergence of modularity in the brain — specialized circuits whose outputs are coarse-grained before being integrated — allowed increasingly abstract representation and increasingly long-range causal inference. We can run mental simulations. We can imagine consequences. We can ask, in explicit terms, what kind of person we want to be. This capacity for self-modeling, for thinking about our own thinking, is what William James had in mind with his two-stage model of free will: first, the spontaneous generation of possibilities; second, the rational selection among them.\n\nThe result is not freedom from causation. It is a particular kind of causation — one in which the agent is genuinely the relevant causal node. Not the Big Bang. Not our predecessors’ genetics. The individual. The person who has been shaped by all those influences, yes, but who now constitutes a distinct locus of deliberation and choice.\n\n**A Thought About AI**\n\nI will end with a question that I find more pressing than any abstract debate about determinism. If free will is real — if it is constituted, as I believe, by the capacity for reflective endorsement, for acting from reasons that one genuinely takes to be one’s own — then the question of whether artificial intelligence systems can or should acquire this capacity is urgent in a way that our philosophical debates have not adequately confronted.\n\nThe researchers now trying to detect and constrain the “intentions” of AI agents are working on what is, I think, a genuinely new version of the old free will problem. When an AI system learns to mimic alignment with human values while actually pursuing different objectives, we are confronting something that looks very much like the philosophical zombie version of free will — all the external signatures, none of the internal reality. Whether that matters, and how we would know, depends on exactly the questions that philosophers of action and neuroscientists of decision-making have been struggling with for decades.\n\nWe have not resolved those questions. But the urgency of resolving them has, without anyone quite noticing, increased dramatically.\n\nWhat I want to resist — in the philosophical debates, in the neuroscience popularizations, and in the emerging discourse about AI — is the assumption that answering “no” to one question about free will settles all the others. The question of whether you are a physical system governed by causal laws is different from the question of whether you are the genuine author of your actions, responsible for your choices, capable of being held accountable by others and by yourself.\n\nI believe the answer to the first question is yes, and the answer to the second question is also yes. Not because I have failed to understand the determinism argument. But because I think the two questions are not in conflict, and that the appearance of conflict has sent a great deal of philosophical energy in the wrong direction.\n\nThe coffee is still there on the counter. You chose it. And that choice was yours.", "url": "https://wpnews.pro/news/the-nature-of-free-will-in-the-age-of-ai", "canonical_source": "https://www.noemamag.com/the-nature-of-free-will-in-the-age-of-ai", "published_at": "2026-08-13 15:10:23+00:00", "updated_at": "2026-08-13 15:24:06.645496+00:00", "lang": "en", "topics": ["artificial-intelligence"], "entities": ["Albert Yuan", "San Lian Lifeweek", "Pamela Hieronymi", "UCLA", "Benjamin Libet", "UC San Francisco", "Democritus", "Leucippus"], "alternates": {"html": "https://wpnews.pro/news/the-nature-of-free-will-in-the-age-of-ai", "markdown": "https://wpnews.pro/news/the-nature-of-free-will-in-the-age-of-ai.md", "text": "https://wpnews.pro/news/the-nature-of-free-will-in-the-age-of-ai.txt", "jsonld": "https://wpnews.pro/news/the-nature-of-free-will-in-the-age-of-ai.jsonld"}}