@Feanor Can we put a lower bound on "significant", maybe by specifying something along the lines "both need to invest x% of personnel/compute/ressources"?

@Primer Ah I'm hesitant to impose constraints already because there are so many ways this can unfold, and i do not want the imposed constraints to rule out initiatives that otherwise would've qualified under the spirit of the market. The team sizes change, and if I had announced a size last week when OpenAI had like <10 members on interpretability vs now when they are announcing superalignment, then I'd have to modify the constraint again. Anyway, it should not be a visiting researcher or a minor project, only major projects will count, and let's say that the websites of the orgs announce it as a joint initiative where both teams share internal resources and personnel, that'd count

@Feanor Sounds good to me. I'd suggest an update to the resolution criteria.

Sibling market for 2030

It seems a safety project collaboration is increasingly likely everyday for a few reasons.

Added 1000 Mana subsidy