Alibaba Cloud founder Wang Jian rejects AGI as Yampolskiy calls superintelligence uncontrollable
alex_verem · x · 2026-09-03
At the AI for Good summit, AI safety coiner Roman Yampolskiy and Alibaba Cloud founder Wang Jian exchanged each other's hardest questions. Yampolskiy asked how Wang plans to control general superintelligence; Wang said he doesn't believe in AGI at all—"AI is AI," a capability that grows over time, and the label mostly makes people nervous. Wang asked how to make AI safe without slowing development; Yampolskiy said his research points to no safety for advanced AI, calling it an impossible problem like a perpetual safety machine, and advocates stopping general superintelligence in favor of narrow, testable tools. One says it can't be controlled; the other says it doesn't exist.
More from AGI Musings
- Comparing the dot-com bubble, blogger predicts AI will collapse to a simpler form — rickasaurus · 2026-09-03
- 3,000 Paying Agents in 10 Weeks: Financial Datasets Sells Data to AI Agents — ryanbed · 2026-09-03
- Researchers debate CoT monitorability: what replaces it when CoT goes away? — LauraRuis · 2026-09-03
- KOL urges OpenAI and Anthropic to flood the world with optimistic AI ads for ordinary people — signulll · 2026-09-03
- If a multimodal AI accurately reports light vs dark, why not claim perpetual inner darkness? — yeastsplainer · 2026-09-03
- Researcher: Ideas are now nearly worthless, careful experiments matter more than ever — cjmaddison · 2026-09-03