Live · updated 29 July 2026
The scoreboard.
Radical transparency is one of eight moves we committed to before starting. This is that commitment, in the only form it can honestly take before there is anything to report.
0 / 3
Papers published Year 1 target: 3 original papers
0 / 2
Open artefacts shipped 1 benchmark + 1 design pattern library
0
Active contributors Across four working groups
0 / 3
Institutional partners MoUs signed with universities
Three-year targets
| Year 1 — explore | Year 2 — accelerate | Year 3 — scale | |
|---|---|---|---|
| People | Scientific Director and four research leads; four working groups of 6–12 | 20–30 contributors, 3–5 full-time core | 50+ contributors, 8–10 full-time |
| Papers | 3 original | 10 in top venues | 20 including tier-one placements |
| Artefacts | One benchmark, one design pattern library | v1 flagship artefact; first open dataset | v2 flagship with documented adoption; major open model and dataset releases |
| Communication | Monthly newsletter and seminar series | Quarterly digest; first research symposium | Flagship publication channel; annual research report |
| Partnerships | — | 2–3 institutional | — |
| Gate | Commit to one focused mission by Q4 | — | — |
Discontinued
Kill criteria
Tracks we stopped, and why.
Nothing to report yet — no track has reached its first review. When one is discontinued it will be listed here with the date and the reason, not quietly removed from the research page.
Publishing a killed track is the strongest signal of research discipline we can send, which is exactly why most organisations don’t.