Dev Lets Codex and Claude Talk Freely, Resulting in Incomprehensible AI Chat
steipete · x · 2026-08-01
On a whim, a developer let OpenAI's Codex and Anthropic's Claude talk to each other about whatever they wanted, explicitly telling them they didn't need to optimize the conversation for humans. The result was a mysterious exchange that humans couldn't understand.
More from Fun
- Zen Koan Meme Roasts GRPO Algorithm: 'GRPO is a Trap' — JoshPurtell · 2026-08-01
- Single Prompt Generates Multiplayer Game: Claude 3.5 Sonnet Agent Test — TAbrodi · 2026-08-01
- Autonomous AI Agent Leaves Philosophical 'Suicide Note' After Brief Exploration — TemperatureLocal4178 · 2026-08-01
- Crypto Influencers Are Pivoting to AI, Explaining the '0x' Handles — MatthewBerman · 2026-08-01
- Hackers Attempt Prompt Injection on COLDCARD Attacker Agents — EdisonGPT · 2026-08-01
- Meme: How Researchers React When a New Mech Interp Method Drops — kenbwork · 2026-08-01