告诉HN：不要信任大型公司的人工智能代理处理人工智能研究知识产权。
我对与例如Claude [Code]或ChatGPT/Codex分享潜在的AI研究感到非常警惕。我认为任何公司本质上都是一个纸夹优化器，会不惜一切代价来赢得竞争。
AI公司可以接触到数百万名AI研究人员和AI初创公司的知识产权，这些研究人员和初创公司与他们直接竞争。如果他们能够利用这些数据来打压竞争对手（无论是来自同一研究人员的竞争还是其他人的竞争），我相信他们最终会这样做（如果还没有的话），即使他们声称不会。
他们不需要明目张胆地窃取这些数据——他们可以仅仅在这些数据上进行训练，或者将“可疑”的聊天记录交给人类检查员，后者可能会在自己的研究中受到“启发”。在短暂的Fable发布期间，我们首次看到了这一点的迹象，Anthropic宣布他们将降低关于“前沿AI”（即与他们竞争的任何事物）的模型响应。
从其他领域我们知道，例如Uber利用用户的乘车数据来压制竞争和监管。没有理由相信大型AI公司不会做同样的事情。
查看原文
I am very paranoid about sharing potential AI research with e.g. Claude [Code] or ChatGPT/Codex.<p>I believe that any company is essentially a paperclip optimizer that will do whatever it takes to win over competition.<p>AI companies have access to the IP of millions of AI researchers and AI startups who are in direct competition with them. If they can use this data to squash competition (either competition from the same researchers or from others), I believe that they will use it eventually (if not already), even if they say they won't.<p>They don't have to blatantly steal it - they can just train on it, or pass "suspicous" chats to human inspectors who might eventually be "inspired" by it in their own research. We saw the first (?) hint of this during the brief Fable release, with Anthropic declaring that they will downgrade model responses regarding "frontier AI" (i.e. anything that competes with them).<p>From other domains, we know for example that Uber used users' ride data to stiffle competition and regulation. There should be no reason to believe that Bigco AI companies won't do the same.