CURTIS — ARCHIVE SCIENCES DIVISION

The research arm.

The unsolved problem in AI is not capability — it is custody. Who holds the record of what your AI did, can you verify it, and will it still open in thirty years? Three studies. Every finding ships in the product line, not a paper cemetery.

CURTIS — RESEARCH LEDGER

WHAT WE STUDY, WHERE IT SHIPS, WHO IT SERVES

thesiscustody is AI's open problem — capability is rented, records are kept

methodfindings land as shippable code — every study has a live artifact

standard.curt v1 — portable, byte-stable, readable anywhere, forever

verificationappend-only receipts, sequenced and checkable — audit as a product

partnersuniversities and archives welcome — [email protected]

Partner with the arm see the product line

STATED PLAINLY

There is no grant funding yet, no endowed lab, no whitepaper paywall. The research arm is one working engineer and an agent fleet, and every result has to justify itself by shipping. What you see is the method: study the custody problem, land the answer in a product, version it.

Funding changes the pace, not the terms — findings stay open, formats stay free, and the records stay yours.