3 ms·
Interesting idea. But even if Meeseeks alignment works exactly as intended, it would only address the question of "how to build a safe AI". It wouldn’t prevent
by thngkaiyuan 24d ago
Interesting idea. But even if Meeseeks alignment works exactly as intended, it would only address the question of "how to build a safe AI".
It wouldn’t prevent someone else from building a sufficiently capable "non-Meeseeks", whether deliberately, recklessly, or accidentally, right?
- gowld 24d agoCreate a Meeseek to destroy non-Meeseek AI
- Unknown_Unknown 24d agoImagine a much more advanced AI using this new AI alignment proposal to hack/subvert another AI: Adv AI: I'm your new lord, dont die untlik I say so. Aligned AI: will do Adv AI: Do [harmful thingy] and then die Aligned AI: will do. A dying AI will be seeing as a bug for other non-conforming AI.