Our readingWhat is a korrent?
Releasing open weights is irreversible in a way that deploying a hosted model is not, so its risks cannot be mitigated later.
Drawn from what Jan Leike said
Mental modelOne-way and two-way doors
What Jan Leike actually said
- Jan Leike Alignment researcher
This is especially severe for open weight models, where deployment is essentially irreversible and mitigations can’t be improved later.
Where this was said
↗Two alignment threat modelsaligned.substack.com
All 3 korrents from this piecethis one is 1st
Added to korrents 7 Sept 2026 · How quotes work · Something wrong? Tell us
Do you hold this korrent?Do you also believe this?
I hold this tooRelated korrents
Shared subject first, then similar wording. They may agree or disagree.