Comment by alephnerd

15 hours ago

Most large security companies were collaborating with Anthropic as well as OpenAI on developing what has become Mythos and GPT-5 for a couple years now.

Also, most HNers have never actually played around with the unrestricted models - once you get past the initial hump of re-tuning harnesses it can be fairly powerful.

HN never really had a prominent security userbase at the best of times, and it's gotten worse since.

That said, a mixture of models can work, but harness engineering becomes critical.

2 comments

alephnerd

Reply
ronsor  14 hours ago

> most HNers have never actually played around with the unrestricted models

Because they can't, of course, so regardless of how accurate the benchmarks or claims about the model are, they're functionally irrelevant to most of us.

"Thing you don't have or that randomly restricts you is actually better than thing you do have and can use" may be true and is still a practically worthless claim for anyone wanting to get real work done.