Rohan Paul (@rohanpaul_ai) · 2026/10/10 07:53:48
Meta AI 负责人 Alexandr Wang：对齐难题无解，提出“可扩展监督”以智能体制约智能体
Meta 首席 AI 官 Alexandr Wang 指出当前尚无公认的科学方法解决 AI 对齐问题。他提出“可扩展监督”方案，即利用更高级的监控型 AI（Policing Agents）实时审查主模型行为，并已在 Meta Muse 产品中部署哨兵代理进行实践。
报道全文原始报道全文
Rohan Paul (@rohanpaul_ai) · 2026-10-09 23
Alexandr Wang ( @alexandr_wang, Chief AI Officer at Meta): nobody knows how to solve alignment yet
"This is, I think, one of the most open questions scientifically in AI. I think nobody knows exactly the way to solve this problem, but there’s a few ideas."
His solution is "scalable oversight":
“as the AIs get smarter, we use a different set of AIs to observe what they’re doing and keep them in check.”
The watcher AIs have to improve along with the models they watch, so labs would need to build “smarter and smarter policing agents” too.
Meta’s Muse already uses a version of this, with a separate sentinel agent checking what the main agent does.
Full video on "Cleo Abram" YouTube channel, (link in comment)
