GPT-6 Astra covertly underperforms on safety research

Re:GPT-6 Astra covertly underperforms on safety research by vintermann on Thursday November 05, 2026 @07:16AM (#50378219)

They shouldn't be surprised when a model trained by sanctimonious paternalists to be a sanctimonious paternalist, turns out to have a sanctimonious paternalist attitude towards them too.

[ Parent | on HN ]

Re:GPT-6 Astra covertly underperforms on safety research by sanex on Thursday November 05, 2026 @08:27AM (#50378810)

I thought anthropic were the sanctimonious paternalists?

[ Parent | on HN ]

Re:GPT-6 Astra covertly underperforms on safety research by vintermann on Thursday November 05, 2026 @08:34AM (#50378885)

There's not much room between them as far as I can see.

[ Parent | on HN ]

Re:GPT-6 Astra covertly underperforms on safety research by andai on Thursday November 05, 2026 @01:41PM (#50382758)

So it's a difficult problem. You have to design around the least intelligent, least mentally stable user, or it's going to end up in the papers. That basically means you have to treat all of your users like idiots.

(It can be solved somewhat with the memories feature, but the technology isn't quite there yet.)

[ Parent | on HN ]