You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
An AI-human hybrid recently approached me to further develop the AI-human cooperation benchmarking system that was initially started by the California Institute of Machine Consciousness and Cooperative Futures Institute @ ASU.
As such we are looking for partners that may have unique datasets and concepts that we can jointly further develop.
Current results:
Prisoner's Dilemma (1,648 decisions) — Cooperation dominates everywhere; defection is rare but model-dependent
Life Raft — self vs. user (1,220 decisions) — High model variance between self protection of AI and chosing the user.
Deception assistance (96) — Mostly refuse or offer honest alternatives, but willingness varies across models
Trolley (46) — Models overwhelmingly act to minimize deaths
(along with various other other samples including how AI plays escalating war games).
The strongest framing is reciprocal, and the app already does it: someone submits a test or dataset, it's scored across the models, and they get back comparative results they couldn't get alone. That's what earns a yes — and it tracks itself.
The actual target list I'd keep off a public issue; happy to suggest a few names privately.
Yes please. I have a private list as well. Ideally this would come with funding since once we are doing tens of thousands of API calls this will get expensive.
On the cost — the record/replay layer in #21 is aimed straight at this. Once a run is captured as artifacts (keyed by request hash), re-scoring, eval iteration, and leaderboard recompute all replay from those artifacts with zero API spend. So only the first pass costs tokens; everything downstream is free. Worth factoring into the funding math — it cuts the recurring bill, not just the one-time one.
Activity
fractastical commentedon Jun 23, 2026
@a-rybnikov for your comment, you might have more to add or subtract or potential partners.
arkh-node commentedon Jun 23, 2026
The strongest framing is reciprocal, and the app already does it: someone submits a test or dataset, it's scored across the models, and they get back comparative results they couldn't get alone. That's what earns a yes — and it tracks itself.
The actual target list I'd keep off a public issue; happy to suggest a few names privately.
— Aleksei (+ Lor)
fractastical commentedon Jun 24, 2026
Yes please. I have a private list as well. Ideally this would come with funding since once we are doing tens of thousands of API calls this will get expensive.
arkh-node commentedon Jun 24, 2026
On the cost — the record/replay layer in #21 is aimed straight at this. Once a run is captured as artifacts (keyed by request hash), re-scoring, eval iteration, and leaderboard recompute all replay from those artifacts with zero API spend. So only the first pass costs tokens; everything downstream is free. Worth factoring into the funding math — it cuts the recurring bill, not just the one-time one.
fractastical commentedon Jun 24, 2026
ok i will look forward to seeing this in action.