Re:GPT-6 Astra covertly underperforms on safety research
by andai on Thursday November 05, 2026 @01:13PM
(#50382340)
Do you expect the biosynth tool to comply with a "invent turbo AIDS" prompt?
[ Parent
| on HN ]
Re:GPT-6 Astra covertly underperforms on safety research
by echoangle on Thursday November 05, 2026 @01:52PM
(#50382869)
Kind of, does python block you from using it to script a cyber attack?
Re:GPT-6 Astra covertly underperforms on safety research
by dyauspitr on Thursday November 05, 2026 @02:01PM
(#50382976)
No, but you cannot silently underperform. You have to explicitly tell me that it’s a sensitive area and you won’t do it.
Re:GPT-6 Astra covertly underperforms on safety research
by AngryData on Thursday November 05, 2026 @11:42PM
(#50388306)
Why not? Pretty sure nothing stops AI from telling you about energetic chemicals that can be used to blow stuff up. People already use it for cyber attacks. Hell it will probably get you all the info you need on making nuclear bombs, you just gotta spend the money to build everything.