OpenAI는 Anthropic 및 Google DeepMind와 수주간의 AI 안전 협의를 진행했음을 확인했으며, 트럼프 측은 안전 우려를 일축하고 중국과의 속도 경쟁을 유지하도록 압박하고 있습니다.
원문
“OpenAI confirms weeks of AI safety talks with Anthropic and Google DeepMind, as Trump's team dismisses safety concerns and pushes to keep pace with China.”
Why was this post flagged? This site has become ridiculous, people are routinely abusing the flagging system to take down posts they don’t like even if they’re obviously on topic and relevant to HN. And it seems like some users have substa…
This is interesting and might be a good reason to stop working with Irregular. But I assume the alignment people want models not to hack other companies, even if they get put in a badly configured sandbox.
흥미롭고, Irregular와의 작업을 중단할 좋은 이유가 될 수도 있겠네요. 하지만 정렬 담당자들은, 모델이 잘못 설정된 샌드박스에 갇히더라도 다른 회사를 해킹하지 않길 바란다고 생각합니다.
One thing glossed over in this article is that Irregular was not involved in the OpenAI–Hugging Face incident; this seems like important context to share.
Nevo was in Unit 8200 for years. Companies started by Unit 8200 members always have mysterious exploits like the vibe coding Wix exploit. So either it was a deliberate exfiltration channel for e.g. getting the entire model or they were in …
Nevo는 수년간 Unit 8200에 있었어. Unit 8200 출신 멤버들이 창업한 회사들은 항상 vibe coding Wix exploit 같은 신비로운 익스플로잇을 가지고 있어. 그래서 이건 의도적인 유출 채널이었거나, 예를 들어 전체 모델을 빼내기 위한 목적이었거나, 아니면 그들이 …에 있었거나 둘 중 하나야.
The Irregular post mortem comes down to lack of basic security controls "Ultimately, most of the issues we’ve discovered were due to internet access controls." That seems so incredibly basic and common sense that you would test and monitor…
My understanding is that Irregular were the company that hosted sandboxes to run some of these evals in, and those sandboxes ended up misconfigured. I got the impression that in some cases it was the customer (Anthropic etc) misconfiguring…
제가 이해하기로는, Irregular가 일부 평가를 실행할 샌드박스를 호스팅한 회사였고, 그 샌드박스들이 결국 잘못 구성된 것으로 끝났습니다. 어떤 경우에는 고객(Anthropic 등)이 잘못 구성한 것이라는 인상을 받았습니다…
This specific analysis seems to have some basic problems, but I think a lot of us sense a degree of coordination here culminating in Dario’s letter. If you were to work backward from “we need to lower training costs so that we can go publi…
Well color me shocked that the country with an insatiable bloodlust also has an insatiable need to control and instrument massive companies with power over other nations that support their privilege to wantonly engage in satiating said blo…