Because of hallucination risk, you should use AI first on topics you already understand

Added to korrents 4 Sept 2026 · How quotes work · Something wrong? Tell us

Where this collides

• Cannot both be true

  Because an individual is an expert in at most one thing, an LLM's broad shallow expertise lets them do things they could not do before, whereas an organisation only gets better at what it already did.

  Karpathy says an LLM broad shallow expertise lets a person do things outside their one field that they could not do before. Mollick says that because of hallucination risk you should use AI first on topics you already understand. The advice and the promise point at exactly opposite parts of the map.

Shared subject first, then similar wording. They may agree or disagree.

• More informed AI insiders often come down with AI psychosis, so one should be very, very careful when speaking with them. TCTyler Cowen31 Aug 2026
• Relying on AI misbehavior appearing gradually during training to catch misalignment may fail at exactly the most dangerous time. ZMZvi Mowshowitz2 Sept 2026
• Language models hallucinate exactly when you least want them to: on important, difficult questions they are at their most confidently wrong. EMElon Musk9 Nov 2023
• AI should be used as much as possible for writing as reporting, but should not replace writing as thinking. TSTara Seshan30 Aug 2026
• Rather than having AI just do the task, people should use it to help build the right mental models. FGFlora Guo25 Mar 2026

On the map

Open the map on its own page