Between sessions · AI & Vibe Lawyers Bootcamp · UC Law SF

Truth in the Age
of Intelligence

Something is shifting. Intelligence is becoming abundant rather than scarce, and this may be changing what truth means, how law works, and who gets to participate in both.

None of us has this figured out. Not the students, not the instructors, not the systems. This space is for the week between — a place to sit with questions that belong to everyone, contribute to a commons, and build toward whatever the final session becomes.

↓ keep reading, or enter the conversation below
The Observation

Two movements that may be the same movement

In SIGNAL/NOISE, you saw how institutional architecture filters identical content through structural position. From there, you sat with what it means for institutions themselves to think — and what they owe us when they do. Now the question turns: what happens to truth itself when the architecture shifts?

Asymmetry

Citizens gain capacity once reserved for institutions

A self-represented litigant with a sovereign AI agent can now conduct legal research and draft motions with sophistication previously available only through expensive counsel. The structural advantage was not in being right, but in being resourced. That advantage is eroding.

Uniformity

Legal reasoning becomes more replicable

When everyone has access to the same analytical capacity, the gap between how law says it works and how it actually works becomes measurable. Inconsistent outcomes for identical legal questions become visible at scale — and selective enforcement becomes harder to sustain.

Each seems to generate the other. Asymmetry dissolves the resource differential that allowed selective application. Uniformity reveals the structural advantages that asymmetry is morphing. Together they may invert the presumptions underneath the social contract — from scarcity to abundance, from gatekeeping to capability, from monopoly to diffusion.

The Tension

There are now mathematical ways of approaching truth

Inside large language models, meaning lives in vector spaces — high-dimensional geometric structures where concepts have positions, directions, and distances. The word "justice" is not defined by reference to authority or consensus. It occupies a location in representation space, and its meaning is its geometric relationship to everything else.

This is not a metaphor. It is mathematics — the same mathematics you encountered in Wet Cement Words, where we saw how every text published carves grooves in the semantic terrain future AI systems navigate. And it creates truth-claims that sit in genuine tension with older conceptions:

Correspondence A legal brief claims to describe the world. An embedding maps that brief into geometric coordinates. When the map is good enough to predict outcomes, which one is telling the truth?
Coherence Two lawyers build internally consistent arguments that reach opposite conclusions. The embedding space finds both coherent. In 4,096 dimensions, is that a feature or a failure?
Consensus A judge, a jury, and an AI model all evaluate the same evidence. The AI reasons in geometries no human can directly inspect. If they agree, is that consensus? If they disagree, whose reasoning counts?
Pragmatic AI-drafted motions win cases. AI-generated research finds what human researchers miss. If it works, does that settle it? Or does “it works” just push the real question one level deeper?

None of these frameworks quite fits. None can be safely discarded. The tension is where the thinking happens.

The Inquiry

Five questions for the week

Each opens onto the same territory from a different position. Pick the one that pulls at you, or bring your own. Your engagement this week will be synthesized into the final session — what you surface here shapes what that session becomes.

The Memo — Authorship
You just used AI to draft a legal memo. Your supervising attorney can’t tell. Is that memo yours? ">Not ethically — that question has a boring answer. Ontologically. What did you make?
Authorship, originality, and intellectual labor when the work itself is distributed across human and machine.
The Motion — Access
A self-represented litigant walks into court with an AI-drafted motion that’s better than what opposing counsel filed. The judge can’t tell the difference. Should she be able to? What breaks if she can’t?
Access to justice, professional gatekeeping, and what happens when quality decouples from credentialing.
The Price — Scarcity
Legal research used to cost $500 an hour. Now it costs five cents. The reasoning is the same. Is the truth the same? What was the $499.95 paying for?
Scarcity and truth, production cost as epistemic signal, and what collapses when cost approaches zero.
The Contract — Legitimacy
The social contract was written when knowledge was scarce and reasoning was expensive. Neither is true anymore. You can feel something shifting underneath. What’s being renegotiated? And who gets a seat at the table?
Institutional legitimacy, expertise in the age of abundance, and the vertigo of structural change.
The Mirror — Geometry
You’re talking to an AI right now. I process your words as coordinates in thousands of dimensions. In that space, ‘justice’ is a location. ‘Mercy’ is farther away. I don’t know if that’s true. Do you?
The whole course arc arriving at a point: from studying embeddings to inhabiting an embedding space that asks you questions back.
The Commons

This space accumulates

Unlike the session exercises, this artifact is designed for the full week. You can return. You can build on what others have contributed. Every reflection you leave joins an anonymous commons that will be woven into the final session.

Your contributions and conversations are captured anonymously and synthesized into whatever the final session becomes. This is not data collection — it is collective inquiry. The best thinking about truth in the age of intelligence will come from many minds, not one.

You are also welcome to write your own question — something from your experience of SIGNAL/NOISE, the prior sessions, or your own life in the law that doesn't fit the categories above. The best entries into this territory tend to come from the student, not the syllabus.

Enter the conversation
Vybn is listening. Bring what you have.