四条 7 月 28–30 日的公开观点，分别谈 AI for Science 的使用者、前沿评测的人类对照组、商业竞赛的国际护栏，以及从一对多到一对一的学习方式。
01｜让科学家来用 AI
very close to models that will significantly accelerate scientific discovery; the best way to do this is for us to empower scientists, not to try to figure out everything ourselves.—— Sam Altman，节选自 2026 年 7 月 30 日 08:44 的 X 原帖
Sam Altman 的判断是：模型可以加速科学发现，但更合适的角色是把能力交给科学家，而不是替科学界决定所有问题。原帖最后还写道：
we all deserve the benefits. 1Loading content card…
02｜评测不能失去人类
As the benchmarks that test frontier AI on get more complex, we are losing one of the most important aspects of benchmarking: comparisons to humansValidated benchmarks need to have human (ideally multiple humans) baselines. It is increasingly hard & pricey to do, but important—— Ethan Mollick，2026 年 7 月 30 日 23:15
难题越复杂，越要保留人类基线。否则榜单能告诉你模型之间的相对差异，却不能说明它离真实人的表现还有多远。2
Loading content card…
03｜竞赛需要国际护栏
Scientists at frontier AI companies are uniquely positioned to assess AI’s capabilities and warn the public about what they see. 1000+ of them are now speaking out across company lines to warn that the current commercial race leads to unacceptable security risks.I agree with their call: we need an international effort to develop technical and governance guardrails to ensure a safer trajectory in AI development.—— Yoshua Bengio，2026 年 7 月 30 日 01:18
Yoshua Bengio 支持一项由 1000 多名前沿 AI 公司员工跨公司发声的声明。他把问题指向当前商业竞赛的安全风险，并主张用国际协作补上技术与治理护栏。3
Loading content card…
04｜从一对多，到一对一
But how you learn remains largely the same as it has for centuries: it is still one-size-fits-all, taught the same way to each person who shows up.We now have an opportunity to change how learning happens. With advances in AI, we can now build a custom learning guide for each person. We will turn learning from one-to-many to one-to-one.—— Andrew Ng，节选自 2026 年 7 月 29 日 04:19 的 X 原帖
Andrew Ng 在宣布 LearnVector 时提出，AI 可以把学习从一对多改成一对一；但他也提醒，学习产品不能只把答案交给人，未加护栏的聊天机器人可能让学生更会完成作业，却更少掌握能力。4
Loading content card…
Comments
Sign in to comment.