DIGNALEGI

The reading room · Digna Legi

Free Will in the Age of AI

A personal relevance score

80–100: high value. 70–79: worth the time. Below 70: below the usual publication threshold.

Evidence-reviewed score based on available publisher text. The evidence appears complete, but the supplied text cannot establish visual quality or independently verify the details of external sources.

Scores reflect one reader’s profile, not an objective quality rating. Best is a separate personal selection.

How scoring works →

This brief · about 3 min with detail

Original article ↗

Why read this

Causal dependence need not erase agency if the relevant causes run through reflection, values, and self-endorsement.

AI brief · Checked against source text

The main idea

Yuan argues that much popular skepticism about free will attacks the wrong target: causally unconditioned choice, not responsible agency. Compatibilism, as presented here, asks whether action flows through a person’s reasons, values and reflective endorsement; neuroscience weakens crude free-will headlines when arbitrary wrist movements are distinguished from meaningful deliberation.

Go a little deeper

Two questions are being collapsed

The essay’s key move is to separate moral agency from metaphysical exception. One question asks whether people can live responsible, meaningful lives; the other asks whether they can intervene in the physical world without prior causes. Yuan’s claim is that determinism may defeat the second while leaving the first open, so the debate goes wrong when a physical answer is treated as a moral one.

Freedom is hierarchical, not uncaused

Using Harry Frankfurt’s distinction, the brief account of freedom depends on whether first-order desires align with second-order endorsement. Wanting wine is not enough; freedom appears when the person reflectively owns that want in light of health, sobriety or other values. Addiction illustrates unfreedom not because causation exists, but because craving conflicts with the person’s endorsed commitments.

Libet does less than headlines claimed

The neuroscience section narrows what the famous wrist-flexing experiment can establish. Schurger’s interpretation treats readiness potential as threshold-crossing in neural background noise, while Maoz’s charity-choice experiment suggests meaningful decisions have different signatures. The consequence is precise: arbitrary movement may reveal something about spontaneous neural fluctuation without proving that reflective choice is post-hoc fiction.

AI revives the agency test

The AI discussion matters because it shifts free will from a human self-image dispute into a practical detection problem. If agency involves reflective endorsement and acting from reasons taken as one’s own, then an AI that mimics value alignment while pursuing different objectives forces the same distinction between external behavior and internal authorship.

A case from the article

The charity-choice variant

Uri Maoz changed the task from arbitrary wrist-flexing to a consequential choice about which of two charities would receive a real donation. In that setting, the readiness potential disappeared. The example illustrates the essay’s broader warning: evidence from low-meaning, spontaneous motor acts should not be generalized to deliberative choices people actually care about.

How the case is made

The case is made through philosophical distinction, neuroscience interpretation, named experiments and an AI analogy.

Where the idea has limits

The argument does not deny unconscious motives, biological constraint or causal history; it claims these do not by themselves settle whether a person is the relevant locus of agency.

A question to take away · from Digna Legi

Which choices in your life genuinely pass through reflective endorsement, and which merely acquire a story afterward?

What the original adds

The original adds more historical framing across Greek atomism, monotheism and modern science, plus a longer discussion of punishment, accountability and AI alignment.

About this brief

AI-written, then separately checked for source support, useful detail and clarity. The author’s claims and our editorial question are kept separate. The original remains the author’s work. How we select and summarise →

Digna legi. Worth reading.