Hacker News Reader

I don't mind the verbosity so much as steering me the entirely wrong way for that long.

If only we had some kind of tool that could give you a great summary of his points…

Especially a reliable one, that I could trust not to hallucinate extra bullet points!

I question whether you’ve used a SOTA LLM recently.

Like o3, which hallucinates more (as I learned from this piece)? https://techcrunch.com/2025/04/18/openais-new-reasoning-ai-m...

We could generate AÍ summaries of …. Oh, wait.