$800 to Bully AI: Startup's Desperate Stress Test for Forgetful Chatbots
Leading AI chatbots lose 30-60% accuracy over long conversations. One startup's paying $800 to bully them into admitting it.
⚡ Key Takeaways
- AI chatbots drop 30-60% accuracy in long conversations, per 2025 ICLR research.
- Memvid pays $800/day to stress-test memory failures in leading models.
- Hallucinations risk real harm in law, healthcare, and corporate settings.
🧠 What's your take on this?
Cast your vote and see what theAIcatchup readers think
Worth sharing?
Get the best AI stories of the week in your inbox — no noise, no spam.
Originally reported by The Guardian - AI
Related Stories
AI Business
Time Series Interviews: 20 Questions That Cut Through the Hype
AI Business
Granola's 'Private by Default' Notes: Open to Anyone with a Link
AI Business
OpenAI's 8-0 Safety Vote That Doomed Its Own Council — While Erotic AI Flourishes
AI Business