# 弗朗索瓦·肖莱谈AI生物安全风险

- 来源：François Chollet (@fchollet)
- 发布时间：2026-08-31 02:25
- AIHOT 分数：36
- AIHOT 链接：https://aihot.virxact.com/items/cmtg5999903rtroj7silexcpq
- 原文链接：https://x.com/fchollet/status/2094129362991976622

## AI 摘要

弗朗索瓦·肖莱认为，2026年AI网络安全革命带来的快速进展可能外溢至生物学领域，导致合成病原体大流行风险。他指出生物学不同于网络安全，缺乏可验证性且依赖人类数据和湿实验等非数字瓶颈，但合成大流行或已在无AI情况下可行，AI可能加速其扩散，人类对此尚未做好准备。

## 正文

Some people worry that the rapid AI progress driving the "AI cybersecurity revolution" of 2026 could soon spill over into biology, potentially leading to extinction-level synthetic pandemics.

While I would not panic, I would take the risk seriously.

Biology is obviously not the same as cybersecurity -- cybersecurity is a fully verifiable domain, where you can synthetically train unbounded capabilities, and therefore where superhuman skill is straightforwardly achievable. Meanwhile biology is not a verifiable domain, so capability remains critically dependent on human-generated data. Plus, biology has critical non-digital bottlenecks, such as the requirement for physical wet-lab experimentation.

However, it is plausible that synthetic pandemics were already feasible without AI, and that AI might lead to the proliferation of these capabilities. That is a real risk, and we are not prepared for it.