New DeepRSI benchmark debuts as Garry Tan calls RSI's Kitty Hawk moment

A new benchmark called DeepRSI has agent run 48-hour ML tasks to measure recursive self-improvement, and YC President Garry Tan says the next 18 months will be RSI's Kitty Hawk moment.

2026-10-13 ~ 2026-10-13 · 2 related posts

• DeepRSI benchmark pushes agents to 48-hour ML tasks to measure recursive self-improvement — garrytan · 2026-10-13
• Garry Tan: next 18 months are the Kitty Hawk moment for recursive self-improvement — daniel_mac8 · 2026-10-13