Hacker News new | ask | show | jobs
by martin-t 122 days ago
> real-time LLM-generated summaries

Please no, LLMs have been shown to introduce biases from training data into the generated "summaries". LLMs should go nowhere public policy.