Topic
3 posts tagged “ai”.
Why LLM errors need to be designed around rather than waited out, what John von Neumann worked out about unreliable components in 1952, and why a decision-only model like Jev fits his framework better than a chat model does.
Why regulation rarely leads to the intended outcome, why centralizing power removes the only check on it, and why the notion that a government, especially the current one, can act responsibly is laughable.
Why LLMs will never be intelligent, why English cannot replace code, and why the labs teaching the world this workflow will not be the ones that own it.
