Clément Canonne

A list of principles put forth by mathematicians, for mathematicians and other researchers, regarding the use of AI in research. "Number #9 will surprise you!" leidendeclaration.ai

Leiden Declaration on Artificial Intelligence and Mathematics leidendeclaration.ai

Take some time (very little) to weigh on this: "Now Is the Time to Give LLMs Access to the ACM Digital Library" docs.google.com/forms/d/e/1F... (not that I want to influence anyone, but my own position is along the lines of "hell no.")

Now is the Time to Give LLMS Access to the ACM Digital Library Why ACM Believes the benefits outweigh the risks in opening up the ACM Digital Library to large language models docs.google.com

Nice second sentence. arxiv.org/abs/2608.113...

Distribution-Free Halfspace Testing with Samples arxiv.org

Our paper on Entropy Equivalence Testing was just accepted to #RANDOM2026. Congratulations to Joy (Qiping) Yang and Yash Pote, who led the charge! arxiv.org/abs/2605.23225

[2605.23225] Entropy Equivalence Testing arxiv.org

So many good points in this post by @nsaphra.bsky.social: only quoting a couple, to encourage you to read the others. "My colleagues and students adopt the writing quirks they read throughout the day, and their own writing becomes more like an LLM’s." nsaphra.net/post/uncanny/

Life on the Uncanny Precipice | Naomi Saphra nsaphra.net

FWIW, here are Terence Tao’s slides at the ICM on maths and AI: teorth.github.io/tao-web/slid... @teorth.bsky.social (I am not endorsing nor criticizing the content, but this is a useful and thoughtful set of points and views in the discussion, from someone who has deeply enga…

teorth.github.io

First preprint by Jiaqi Mao, currently Masters of Computer Science student at #USyd doing a research project with me (and soon PhD student): arxiv.org/abs/2608.19538

A simple and practical $o(\sqrt{n})$-time algorithm for shortest paths in power law graphs arxiv.org

(this program, by the way, was definitely far from free for Google, but great PR for them, and the TCS community obliged.) research.google/blog/gemini-...

Gemini-backed Paper Assistant Tool provides automated feedback for theoretical computer scientists at STOC 2026 research.google

Preprint: arxiv.org/abs/2510.181... (the paper being in-press at TMLR) OpenReview: openreview.net/forum?id=yIA...

NashPG: A Policy Gradient Method with Iteratively Refined Regularization for Finding Nash Equilibria arxiv.org

Preprint: arxiv.org/abs/2510.181... (the paper being in-press at TMLR) OpenReview: openreview.net/forum?id=yIA...

Verifying your browser | OpenReview openreview.net

Me, reading some badly written one-shotted AI "preprint" claiming to solve an open problem in the least interesting way possible

The OpenAI breakthrough on the unit distance problem seems genuinely impressive to my semi-layman eye. (It is!) One possible takeaway, though, is that if you throw an incredible amount of money and resources at focused research, there WILL be impressive progress. I wish we tried that, too.

The "paperclip maximizer" is a silly thought experiment: an entity mistakes an (imperfect) metric for an absolute goal, and proceeds to diverting all resources to producing paperclips. Silly. In other news, Anthropic and OpenAI are burning millions in a race to establish Erdős' 9173439 conjectures.

I uploaded a picture of a random root vegetable and asked ChatGPT to identify it. It told me it was a stochastic carrot.

FWIW, since reviews for a prominent Theoretical CS conference are now out: if your contribution to a PC is a bunch of verbatim AI-generated reviews and basically nothing else, I believe you should never be invited to a Program Committee again. This isn't even about AI, this is about doing your job.

If you think that AI reviews in conferences "to save time" or "be thorough" or "handle the large # of submissions coming due to AI", ask yourself how much it'll cost, and who will pay. If not "gifted" by companies, this is going to be north of $30+/paper. You expect 1000+ submissions? Budget that.

Related to my previous post about people using GenAI/LLMs for everything, regardless of whether they should, whether it makes sense, and whether what they are doing with it even is something that is worth doing in the first place.

"AI will replace U, they said" "Nothing yoai can do aboait it. It's the faitaire."

@henryyuen.bsky.social I have heard you commented on the recent announcement by OpenAI on X, about one of the 10 problems very, very close to your own research. Would you be OK discussing it here as well?

The Pope taking such a strong stance on AI feels rather unorthodox