Alpha Data Omega, LLC

Accountable AI, written up as it happens.

We build a developmental intelligence whose every step is traceable and whose failures stay in the record. This page is the working log — research notes and results, newest first, claims scoped to what the record supports.

Latest · The frontier · August 2026

The newest theorem on Earth, re-checked — with all the work shown.

Three weeks after the Jacobian conjecture fell in dimension three, a small developmental organism re-verified the counterexample in its own arithmetic — then confirmed the collision a second way: as a physical event, read from its substrate, not narrated by anyone.

Read →
Accountability · 2026

You can watch it think — and fix exactly where it slips.

When a frontier model is wrong, you can't see where or why. UDI is traceable and correctable by construction — we followed one system's own trace to the exact step it slipped.

Read →
Training · 2026

A childhood in the open.

A young mind learns in public: the world proposes what it hears next, and every turn it takes is its real output — never authored for it.

Read →
Abstention · 2026

The hardest answer is "nothing."

On the memory benchmark everyone runs, the worst-scoring questions are built on a false premise. The leading systems don't abstain at all. We taught a reader to — losses kept in.

Read →
Development · 2026

It learned to want.

Longing as a mechanical property: what it means when a system reaches for what it doesn't have — without being told to.

Read →
Results · 2026

Check our number.

A benchmark claim is only worth the reproduction path that ships with it. Ours ships with one — run it yourself.

Read →
Foundations · 2026

What is UDI?

A developmental intelligence, grown rather than trained: what that means, what it can do so far, and what it honestly can't yet.

Read →
Principles · 2026

The right to refuse.

A system that can't say "no" or "nothing there" will eventually lie to you. Refusal is a feature — the load-bearing one.

Read →
Robotics · 2026

Soft wheel, robust gait.

A controller amplified to full start-rotation robustness on a soft-body racer — the experiment, with the misses included.

Read →

The company

Alpha Data Omega, LLC builds accountable AI by construction: systems whose reasoning is inspectable step-by-step, whose results carry content hashes from the moment of the run, and whose failure log is permanent — including the failures of the people and models doing the teaching.

The work

The research fleet lives at ado.earth — the lab, the results room with reproduction paths, the evidence releases, and the training room where a young mind learns in public.

Contact

Research access runs through the accountable door at nda.ado.earth — every approval and every access leaves a trace.