Google DeepMind's New Robot Model ER2 Breaks Hardware in Real-World Test
kamathsblog · x · 2026-08-02
A developer shared a real-world failure experienced with Google DeepMind's newest robot model, ER2. During testing, ER2 provided incorrect physical manipulation instructions, directly leading to broken hardware. In contrast, Anthropic's Claude Opus model successfully provided the correct instructions twice under the same conditions.
More from Fun
- Dev Jokes About Switching Side Projects Weekly from Codex's POV — haydendevs · 2026-08-02
- 40 Engineers Gather for a Jeopardy-Style LSAT Logic Tournament — emilyzsh · 2026-08-02
- Imagining '2001: A Space Odyssey' Remade with AI in 2026 — xiaosun86 · 2026-08-02
- AI Will Solve the Riemann Hypothesis Before Doing Our Taxes — rickasaurus · 2026-08-02
- Gary Marcus Slams AI Community for Desperate 'Astra is ASI' Hype — GaryMarcus · 2026-08-02
- Google Earth Pulls AI Image Feature in Under 24 Hours After Deepfake Disasters — jonerp · 2026-08-02