Article

Hallucinations in code are the least dangerous form of LLM mistakes

Simon Willison · simonw.substack.com

A master practitioner reframes the real risk: errors that look plausible and are never checked.

Open on simonw.substack.com ↗

The question it answers

  1. How do I fact-check AI answers and avoid hallucinations?

Read with it

  1. OpenAI's own research on why models bluff: training rewards guessing over admitting uncertainty. openai.com Article Why language models hallucinate OpenAI
  2. Video Why Large Language Models Hallucinate Martin Keen, IBM Technology
  3. Video OpenAI Explains: Why Language Models Hallucinate & How to Reduce Them YouTube AI explainer
  4. Researchers debate what the hallucination paper actually means for everyday users. ibm.com Podcast Why language models hallucinate (Mixture of Experts) IBM Mixture of Experts
  5. Video How to Fact-Check ChatGPT and Other AI Tools YouTube tutorial
  6. 1,700+ real court cases where fake AI citations got people sanctioned; the best cautionary tale on the internet. damiencharlotin.com Tool AI Hallucination Cases Database Damien Charlotin
eChai Partner Brands The partnership →