link: https://x.com/teortaxesTex/status/2099575222512836893

whole text: https://www.reddit.com/r/NewsWithJingjing/comments/1whrnoi/deepseek_kernel_engineer_shengyu_liu_刘胜与_on_why/pa51kab/

I do not trust Anthropic or OpenAI to do that. In particular, I do not want Anthropic to control the world’s most advanced artificial intelligence or AGI. To put it dramatically, I think the stakes would be comparable to Hitler obtaining the atomic bomb before the Allies did.