Fenner Tanswell

The video of my talk at the Isaac Newton Institute a few weeks ago is now online: "Can Mathematics Be Hacked? Infrastructure, "Artificial Intelligence" and the Cybersecurity of Mathematical Knowledge" www.youtube.com/watch?v=O11k...

A couple of weeks ago, a piece of slop I created went viral (well, a little bit) ... now I've followed it up with a new paper: "What is AI Slop? An Explication" - open access at www.tandfonline.com/doi/full/10.... - feel free to circulate! #philsky #AIslop

New paper out today. I argue that various uses of LLMs by academics can exacerbate existing patterns of marginalisation. I focus on Philosophy, but I'd expect this generalises to other disciplines too. link.springer.com/article/10.1... #PhilSky #PhilTech #LLMs

The JMB special issue that Yacin Hamami, Silvia De Toffoli, Pablo Mejía Ramos and I have been editing is now complete! It explores the relationship between mathematics education and the philosophy of mathematical practice. Explore here: https://www.sciencedirect.com/special-issue/10M3PCCWM15

"This is why we need philosophers of mathematics!" cameroncounts.wordpress.com/2026/10/15/r...

arxiv.org/pdf/2510.20956 Interesting research about LLMs getting more inclined to "self-jailbreak" after being trained in maths and coding.

My biggest regret from this paper is that we didn't take more explicit aim at the exclusionary definitions of CE floating around academic.oup.com/pq/article/7...

it would be really helpful if they separated out the ones that have lean certs from the rest. i think it's just the ones named in this yaml file? github.com/openai/math/...

@fennert.bsky.social talking about this problem just yesterday bsky.app/profile/fenn...

This great article by Tanswell covers exactly this: bsky.app/profile/did:...

The youtube livestream for this will be found here: www.youtube.com/@SorinBangu/...

My short viewpoint article "The Enshittification of Mathematics" is out now in the Mathematical Intelligencer! link.springer.com/article/10.1... Gift link here: rdcu.be/5uYGc2uo454a Check it out and share it with anyone who might be interested! 💩🥳

I've posted some information at jdh.hamkins.org/philosophy-a...

My short viewpoint article "The Enshittification of Mathematics" is out now in the Mathematical Intelligencer! link.springer.com/article/10.1... Gift link here: rdcu.be/5uYGc2uo454a Check it out and share it with anyone who might be interested! 💩🥳

There has been a huge amount of experimental philosophy research on the concept of KNOWLEDGE This research finds that the ordinary concept doesn’t work at all the way philosophers said it does A thread with some examples: 🧵

This take, which is more interesting than my own, came out of a discussion last night about my attempts to describe my views on purpose of mathematical research: terrytao.wordpress.com/2026/09/14/w...

sitting on loads of counterexamples to major conjectures but honestly I'm not in it for the glory, might leak via various channels that won't be traceable back to me just in case other people would find the results useful

For anyone interested, I talk about this in the discussion section of a forthcoming xphi paper on trust and AI philpapers.org/rec/LANTIA-8

It is indeed a huge amount of work and things can be made more efficient by introducing luck. For many grants a simple lottery is efficient. And, if that's too big of a step, asking for abbridged proposals, funding top 10%/removing bottom 50%, then a lottery on the remaining 40%, works too

www.lawfaremedia.org/article/ais-... A very nice article linking LLMs to the long history of stories where your wishes don't always have their intended effect.

Mein erster Gedanke war: "Ach, die Dino-Bots treten auch zur Abgeordnetenhauswahl an? Dabei haben wir für Dinosaurier-Politik doch schon die CDU." #berlinwahl

I thought this was a good time to mention (again) that @fennert.bsky.social and I wrote a paper arguing that using LLMs to generate and verify proofs does not necessarily give us mathematical knowledge as that notion has traditionally been understood: www.mxphi.com/all-issues/v...

My paper on my solution to the rule-following paradox is out in Ergo: doi.org/10.3998/ergo...

"The process is deliberately unhurried, but we will provide updates." www.claymath.org/news/navier-...

According to my memory from last night and the wayback machine, OpenAI updated the Navier-Stokes pdf at Tue, 08 Sep 2026 19:09:35 GMT. The original version was one page shorter and did not contain any citations to Diego Córdoba and Luis Martínez-Zoroa: web.archive.org/web/20260000...

The one upside of this all is that the comments section validated my post's main points repeatedly and with surprising variety.

Major result in formalisation breaking first on a coffee shop's instagram is a new one! www.instagram.com/p/Dc32GtFtzak

Kevin Buzzard's thoughts on Fermat's Last Theorem being autoformalized in Lean. xenaproject.wordpress.com/2026/09/04/f...

I wrote a blog post about how experimental philosophy should address the most toxic aspect of analytic philosophy, namely its insistence that any alternative way of doing philosophy is pointless, not rigorous, or not philosophy ethanlandes.com/posts/2026/0...

I'm getting to teach a seminar on Xphi this autumn. 🥳 What are your favourite Xphi papers? Classics or recent ones appreciated. The more diverse (topics/methods) a selection the better.

I'm getting to teach a seminar on Xphi this autumn. 🥳 What are your favourite Xphi papers? Classics or recent ones appreciated. The more diverse (topics/methods) a selection the better.

Want to learn more about Feminist Logic? Interested in starting a reading group but not sure where to start? Look no further! The DRL is home to a wealth of reading group blueprints and here, we are spotlighting our blueprint on Feminist Logic! 1/2 #philsky #philosophy #DRL #backtoschool

The internet reacts to an improvement in matrix multiplication efficiency from 2.371339 to 2.371177.