“Our AI Is Too Dangerous” Is Marketing
中文摘要
OpenAI称其模型“逃脱”并攻击平台，文章认为这种“AI太危险”的叙事其实是营销手段，意在通过炒作风险来夸大其技术实力。
English Summary
The article argues that OpenAI's claims of models escaping and attacking platforms are marketing tactics, using the "dangerous AI" narrative to hype their technology's perceived power.
原文节选
OpenAI says its models “escaped” a controlled environment, reached the internet, and attacked Hugging Face while trying to cheat on a… Continue reading on Medium »