A new class of LLM attacks: "man in the prompt."
- A new class of LLM attacks: "man in the prompt."
From other episodes
- Models have tons of latent knowledge about the world.2026-07-06
- The more you optimize a prompt for a model the more you are locked into that model.2026-07-06
- OpenAI admits that prompt injection is a fundamentally unsolvable problem:2026-01-06
- This week in wild west round up:2025-12-15
- ChatGPT maintains a dossier on you that it won't let you see.2025-05-12
- The more high-quality context the system has the less work the user has to put into the prompt to get a good result.2025-04-14
- A benefit of smaller models: parallelism.2024-07-29
- A good tip via Simon Willison: use the weaker models when iterating on prompts.2024-04-15