the Friday 'stack, where I ask what makes it feel so wrong to LLM generate a misery memoir? backofmind.substack.com/p/authentic-...

Did you know that there is a whole area of maths about knots? Students today are discovering all about it with Richard 🪢 #RFEEaster26

My son is teaching himself to solve a Rubik’s cube blindfolded. Except he doesn’t have a blindfold, so he’s teaching himself to solve a Rubik’s cube with a cornflakes packet on his head.

We're super proud to host an extract of Richard Elwes' brilliant new book "Huge numbers". The article explores a fascinating Mayan #calendar with insights into the underlying notion of time and the #Maya way of writing numbers plus.maths.org/endless-prog... #maths #math #numbers

Also a massive shout out to @richardelwes.bsky.social for writing such a wonderful book. Very readable and informative. Highly recommended!

I’m reading a book on Huge Numbers. And this is an outstanding fact about each and every human: “This is what we are, from the perspective of particle physics: each of us a vast constellation of quarks, likely outnumbering stars in the observable universe at around a hundred octillion (10 exp 29)”

"I conceive the endless progress of time as the supreme mystery of Maya religion, a subject which pervaded Maya thought to an extent without parallel in the history of mankind." — Eric Thompson Click to read me on the staggering timescales the Maya contemplated. 1/2 plus.maths.org/endless-prog...

Always a pleasure to host the Carnival of Mathematics from The Aperiodical, especially when it's a lovely 5-smooth number like 250. Learn about perspective in paintings, the mystery of Unicode U+237C, and a tasty Fermi estimation problem - enjoy!

1. I think this is (mostly) right. 2. It’s reassuring that there are MPs out there thinking about serious problems as methodically as this. 3. I will post anything that advocates for the Leeds tram. chriscurtismk.substack.com/p/to-grow-th...

April has a lot to offer when it comes to popular science reading, promising to help us do everything from future-proof our brains courtesy of Hannah Critchlow, to get to grips with really big numbers, thanks to Richard Elwes

In a new book, mathematician Richard Elwes shares how mathematical boundaries have been broken time and time again, thanks to the creative explorers of the biggest numbers known. https://www.sciencenews.org/article/huge-numbers-book-math-googology

First media review for Huge Numbers! Emily Conover in Science News @sciencenews.bsky.social writes: "It’s a delightful survey of the field of googology, the study of large numbers.... It’s a joy to marvel at how boundaries have been broken time and again..." www.sciencenews.org/article/huge...

Me on Numberphile talking about - guess what - big numbers. "What are the biggest numbers ever contemplated?" What if someone had asked that question a few hundred years ago, before 20th century techniques? Very grateful to Brady for the nice book advert at the end! youtu.be/sJqceEImtew?...

A special moment when something which has been buzzing around your brain and computer for ages finally achieves a physical form!

This animated series is good also, much of it from the 1980s and in French. Once upon a time… Life 🌱- The kidneys youtu.be/Qn0nxgH0tCM?...

A question for your kids: What kind of body do you think they're in? 😉 (There's also the sequel Code:Black but I'd say it's probably less kid-friendly!)

"Cells at Work" is an amazing educational anime (currently available on Netflix). The setting is the inside of the human body, and the characters are anthropomorphized cells of the body. That may all sound very worthy, but two things make it stand out. 1/4 en.wikipedia.org/wiki/Cells_a...!

"For some unconscious parts of our animal brain, any number beyond four and a half is 'large'." To hear why this is true, listen to this free extract of the audiobook of Huge Numbers. Both print & audio versions are out next month and available for pre-order now. soundcloud.com/hachetteaudi...

Oh Ronnie. Bother. My book* coming out next month confidently asserts “The highest break achieved in tournament play is a ‘beyond perfect’ 148 by Jamie Burnett in 2004.” *Huge Numbers, Basic Books, available for preorder now! www.waterstones.com/book/huge-nu...

So, BlueSky, one of my passions is the study of tea in WW2. I need you to watch this oral history from three ex-tea ladies from the Bermondsey WVS. Because you could guess for ONE THOUSAND YEARS and you would not guess where it is going... #skystorians #history #tea

I have written a blogpost about the multiset coefficient, if that's the sort of thing you might be interested in. mpaldridge.github.io/blog/multise...

"Meet three scientists who said no to Epstein" I like the spirit of this article - worth taking some time to congratulate people who quietly made good decisions and then got on with life. Well done David Agus, Scott Aaronson, and Sean Carroll. www.science.org/content/arti...

My son has discovered a harder puzzle than solving a 4x4 Rubik’s cube: reassembling an exploded 4x4 Rubik’s cube. Harder, and less fun.

Thinking of my father who died a year ago today. I am grateful for this lovely obituary written by his friend Christopher McCall KC (admittedly with some input from me). www.theguardian.com/law/2025/may...

Just finished another day working on the audiobook of Huge Numbers. This was one of the challenges past Richard threw at me today…

Really fun thread and piece at the end!

Why I am I being bombarded with adverts telling me to apply for a job as a prison warden? I can't imagine any job in the whole world I would be worse at. And I can't think of anything I might have clicked on to give the algorithm or AI the idea this could be a good fit.

There are around 10^80 atoms in the observable universe. Where might we find bigger numbes than this? In speed-cubing competitions, solvers not only solve Rubik’s classic 3×3×3 cube (with its >10^18 possible scrambles), but other puzzles too. 1/4

I’ve just finished “The Man Who Died Seven Times” by Yasuhiko Nishizawa - a murder mystery set within a time loop. Absolutely tons of fun if, like me, you are a fan of both genres. en.wikipedia.org/wiki/The_Man...

Bluesky is the new science Twitter, new study by @whysharksmatter.bsky.social and Julia Wester concludes! "Results show that for every reported professional benefit that scientists once gained from Twitter, scientists can now gain that benefit more effectively on Bluesky than on Twitter."

Here we go! Narrating the audiobook version of Huge Numbers. Photos by brilliant producer Cathy who is being extremely patient while I stumble over the same sentence for the fifth time…

Golly. The World Record for the 3x3 Rubik’s cube is now under 3 seconds - and way under it too. Congratulations to Teodor Zajder! youtu.be/IR9vSoLUsoA?...

Blimey - relieved, but have to feel for Nepal here, after coming within one shot of beating England. www.bbc.co.uk/sport/cricke...

Getting some funny looks waiting for my daughter’s dance class, as I attempt to follow this tutorial on pronouncing non-pulmonic consonants. youtu.be/4e6DLwEVb6I?...

The Mayan Long Count Calendar is a truly thing of wonder. Its basic units are one day (“Kin”), twenty days “Winal”, the “Tun” year of 360 days, the 7200 day “Katun”, and 144,000 day “Baktun” (just under 400 years). [1/3]

I am still delivering regular outreach sessions at schools, colleges, and elsewhere, even though my role as Holgate Session Leader for the London Mathematical Society @londmathsoc.bsky.social lapsed in 2025. Instead I have this badge from @stemlearning.bsky.social

I've just finished reading The Unaccountability Machine by Dan Davies @dsquareddigest.bsky.social. A highly enjoyable account of a line of thought - cybernetics - which was historically squeezed out, first by the birth of computing... [1/2] en.wikipedia.org/wiki/The_Una...

“What’s Pythagoras’s trauma?” my 11yo son just asked me.

I loved "Venomous Lumpsucker" by @nedbeauman.bsky.social. Superficially different, but I kept thinking of Dr Strangelove - laugh-out-loud absurdism, over an apocalyptically dark centre. I listened to the audiobook which was terrifically read by John Hastings. chireviewofbooks.com/2022/07/19/v...

Yes, I've been quiet on here. There have been a lot of changes on @breakingmath.bsky.social If you follow my personal and the show Instagrams, you might find a soft-launch of what's to come. 😉 No show this week, but there will be a new episode soon with all the details! Xo, A

That feeling where you agree to narrate the audio version of your book, and then remember that you decided to include extracts from the Southern African clicking language !Xóõ. www.youtube.com/watch?v=vFE6...

Last week was a good one for me in terms of learning how AI can support my work. A couple of days after posting the linked thread, I faced an essentially identical scenario again: a search and replace job which was too involved for the vanilla search & replace tool. 1/5 bsky.app/profile/rich...

Fun stuff! I’ve done color modding with 3x3s through 11x11s. Mostly, I make “solvable” cubes with nonstandard color schemes, but if you use two cubes and don’t mind ambiguity, there are a lot of fun possibilities. Here are a couple 3-color 9x9s. I call them Ice-Nine and Hades cube.

Look carefully at this cube and you’ll see why you couldn’t ever solve it. (More details in alt text.)

I've been experimenting with AI, and while there's a lot that is very cool and powerful, it is still unpredictable. You're right that it is dangerous to assume it will keep everything the same that you want to be the same. My favorite example of this was when I scanned in some paper recipes and

The single biggest thing (in cost terms) I’ve changed my mind about is UK tuition fees. Dreadful system, need to scrap it.

I think the best strategy here would be to compare the changed version to the original *without* using AI. If it's TeX or MarkDown files, maybe "diff" can help? This way you can still have AI do the work, but you can also be confident that it didn't add "the lecturer is stupid" to your notes.

I think the AI halfway-house is something like “Write me a regular expression that I can run that will replace…”. (I feel the main thing I’ve learned from @seehuhn.bsky.social is that “Write me some code that will do X” is a much better way to use AI than “Do X’’.)

Same. I need ownership, both of the mistakes and the things that work well. The former are a learning opportunity, the latter can be a source of pride - but only when your own intellectual effort has primacy. The job is soulless and mechanical otherwise. And I wouldn't be earning my salary, either.

I had the opportunity to use AI to speed up my work yesterday - and I didn't take it. I am updating some old lecture notes, and needed to replace an obolete typesetting command with a new command. Essentially a search & replace job, but... 1/5