Clément Canonne

Reminder/plug: my graduate-level monograph on "Topics and Techniques in Distribution Testing" (FnT Comm. and Inf Theory, 2022). 📖 ccanonne.github.io/survey-topic... [Latest draft+exercise solns, free] 📗 nowpublishers.com/article/Deta... [Official pub] 📝 github.com/ccanonne/sur... [LaTeX source]

We’re hiring in Cambridge CS! Assistant/Associate Professor position in AI and machine learning, including learning theory and related areas of theoretical computer science. Apply here: www.cam.ac.uk/jobs/univers...

Now's a good time to repost this. Please share and engage. #TCSSky cstheory.meta.stackexchange.com/questions/31...

"If we wish to reduce the phenomenon of AI slop flag planting, the main thing we need is for the alternative strategy (publishing lots of low-effort papers) to further lose out for young people." sophie.huiberts.me/blog/2026/me... by @sophie.huiberts.me on the "top 3-paper-only" hiring policy.

Also, I don't know if there is any way to be clearer, but: If you are a theoretical computer scientist or mathematician currently working for OpenAI, you are not one of the good guys.

In more personal news, I've just picked up my new passport. An emutional moment! 🇦🇺

Now's a good time to repost this. Please share and engage. #TCSSky cstheory.meta.stackexchange.com/questions/31...

An internal model at OpenAI has produced a claimed refutation of ETH! Russell Impagliazzo, Mohan Paturi, and I have been asked to help organize an open and independent examination of this claim. 1/4

Also, I don't know if there is any way to be clearer, but: If you are a theoretical computer scientist or mathematician currently working for OpenAI, you are not one of the good guys.

Why is it called entomology and not mothematics

Are you a PhD student in quantum computing and interested in working with our team in 2027 as a student researcher (i.e., an internship that's ~3 months and in-person at our office)? If so, apply here: www.google.com/about/career...

El Niño/La Niña famously cannot be reliably predicted more than ~6 months in advance due to the "predictability barrier": this is a chaotic system! New research from Australia (CSIRO/UMelb) just showed it *can* sometimes be predicted much earlier, up to a full year! www.science.org/doi/10.1126/...

"The research published today in Science Advances has found that the state of the Pacific Ocean at the beginning of a forecast plays a critical, and often overlooked, role in determining whether long-range predictions of ENSO will be accurate." www.eurekalert.org/news-release...

I know the situation with PHP is incredibly worrying, but please shed a tear for ASP/ASP.NET as well. You can hear it, in its corner, alone and shivering, whispering— —"does anyone care? Can anyone check? I have vulnerabilities, too!"

El Niño/La Niña famously cannot be reliably predicted more than ~6 months in advance due to the "predictability barrier": this is a chaotic system! New research from Australia (CSIRO/UMelb) just showed it *can* sometimes be predicted much earlier, up to a full year! www.science.org/doi/10.1126/...

Using initial states to identify successful El Niño–Southern Oscillation forecasts 1 year in advance - ScienceAdvances www.science.org/doi/10.1126/...

Two people cold-emailed me today about an open problem I proposed at a conference in 2024. Both attached what is without any doubt the verbatim output of an AI model, asking for my feedback on it. tl;dr: this is not OK. I will answer your email, but I will not spend time if you yourself haven't.

Do you have high quality work on the foundations of responsible computing, writ large? This covers everything from work on differential privacy to algorithmic fairness to designing mechanisms and markets to AI alignment. (Submit to FORC!) 1/ responsiblecomputing.org/forc-2027-ca...

🎉🎉 Lecturer Liyi Zhou ( @lzhou1110.bsky.social ) from our School was awarded Teacher of the Year for 2026! www.linkedin.com/posts/supra-... Congrats Liyi!

TIL that Heisenberg did his PhD on fluid dynamics, specifically turbulence. That turned out to be really hard, so later on he switched to something easier, quantum physics

🌏 NeurIPS 2026 is coming to Sydney! Connect with the global AI & ML research community across Australasia: 🇳🇿 AJCAI 2026 | Nov 27–30 | Auckland 🇦🇺 ACML 2026 | Dec 1–4 | Melbourne 🇦🇺 NeurIPS 2026 | Dec 6–12 | Sydney Three conferences, one vibrant research community! buff.ly/oWgMVD8

So I am going to show this with an elementary example. This is a thread and I am essentially thinking on the fly. When you are proving a lean theorem you are showing that `(LeanTT + 3 core axioms of lean + your defs + already proven theorems) ⊢ your theorem` live.lean-lang.org#codez=HYQwtg...

The Zoom link for Rachel Zhang's talk, "Linear Time Encodable Quantum Error-Correcting Codes," is now available on our website. See you tomorrow, 1pm ET! www.tcsplus.org/welcome/next...

Also implies something about LLM use in reviewing. First it's just for finding bugs in papers. Then people start trusting the LLM more, and pretty soon the person basically isn't reviewing at all, just rubber-stamping an LLM's analysis of a paper they don't understand.

(you still have to accept *my* papers, obviously)

Proposal for all first-order logic automated theorem provers to only get updated once a year, on January 1. File that under "New Year Resolutions"

U. Colorado Boulder Mathematics is hiring for two tenure-track faculty positions: 1 open area jobs.colorado.edu/jobs/JobDeta... due Nov 15 1 at the interface of AI + math jobs.colorado.edu/jobs/JobDeta... due Nov 11 Please help spread the word! #MathSky #AcademicSky

This is this week! Register here (free) for Rachel's talk: forms.gle/7JyHgQbFSKiA...

Those were simpler times