![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
The Leiden Declaration on Artificial Intelligence and Mathematics
Jun. 4th, 2026 08:34 pm![[syndicated profile]](https://www.dreamwidth.org/img/silk/identity/feed.png){kind=small}
quomodocumque_feedA well-thought-out statement. One thing I like about it is that it is not in any way anti-AI; the declaration merely says that, as with any other piece of technology, we should use it in ways that serve science, scientists, and humanity generally, and beware of using it in ways that are bad for those things. An AI optimist like me should surely sign on, because those bad futures can be, at the same time, unlikely (that’s what I’m optimistic about) but likely enough to be worth worrying about and taking steps now to avoid. If you agree, please consider signing on as an endorser.
People I respect a lot and who I know to have thought a lot about these questions were members of the working group that put this together; hammering out documents like this is a valuable kind of work which I personally hate doing, so I truly appreciate their efforts. Signatories at the moment include bigshots like Peter Scholze, Terry Tao, Ulrike Tillman, Kevin Buzzard, Scott Aaronson, etc. — people who have a wide range of opinions and approaches to the intersection of technology and pure mathematics.
Here’s the comment I included with my signature:
“The declaration represents a clear statement of principles that I hope almost all mathematicians can join in endorsing. Many mathematicians, myself among them, are optimistic about the role artificial intelligence will play in mathematics research. But optimism is no reason to ignore potential negative consequences of these developments. Most importantly, we mustn’t lose sight of our fundamental goal, which is to enlarge and enrich human understanding of the mathematical universe. Our community, over many hundreds of years of working together, has developed a rough consensus about what it is we are actually trying to do. We don’t spend much time explicitly talking about these values; now is the time to do so.”
An especially important point in the declaration is its recognition of “the risk that research questions may come to be prioritized because of their amenability to automated mathematics, rather than expert judgment of their deeper significance.” It is natural for any community to shift its values over time. But if we do it, I think we should do it on purpose and with our eyes open.
![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif){kind=small}
