Cipher_Challenge (cipher-challenge on Bluesky)

@RealGeoffBarton I suppose DJ would be an an easier way to get people to hear you without being able to see your lips move.

Poignant. As a few people know, my childhood ambition (before radio DJ) was ventriloquist ....

We've a few tickets spare for the National Cipher Challenge event at Bletchley on 4th March. With speakers including Rob Eastaway.If you or your school took part in the competition in the Autumn, email us at cipher@soton.ac.uk with your details so we can see if we can fit you in.

@charlesarthur @graceyldn They install mini black holes in the machines to keep them stable - it's why they are so hard to move. That together with the fast spin plays havoc with the measurement of time.

@graceyldn It is MADDENING. Mine does "one-hour" washes that last for any time between 45 mins and 3 hours. Always with 15 minutes left for any length of time.

@jamestitcomb Did the department misunderstand the minister's request and ask an AI to write a skills website perhaps?

The Government’s AI skills website is going well

@FoggyCrazyTimes @rastokke Maybe get him to look at cipherchallenge.org: a story, a mystery, and some tricky ciphers to solve. We have published the solutions to this year's competition, but he can still try them, and we will launch the 25th edition in September.

@rastokke Ohhh I loved math! Puzzles! Solving! But my oldest thinks abstractly and mathematically but hated formal math. I dropped him down a grade. Still hated it. Pulled out logic and puzzles, love! And then we went story mode, BINGO! Now he loves math

1/3💡The problem with "productive" struggle. For kids good at math struggle can be fun because they're successful at the end. For many others, struggle is associated with failure, because they're not successful at the end, and it reinforces for them that they can't do math.

@C_Hendrick "The question is what education looks like when the technology underpinning it improves exponentially" - If one horse can pull a heavy cart at 1 mile an hour and two horses can pull it at 2 miles an hour then think how fast 20 horses could pull it!

#aHR0cHM6Ly9zdWJzdGFjay5jb20vaG9tZS9wb3N0L3AtMTc3NTU2NjYwP3NvdXJjZT1xdWV1ZQ==" target="_blank" rel="nofollow noopener">rynomad.github.io/stubsack/?utm_…

@edujdw @ReemKhan_07 You shouldn't continue to prop up this thread.

@ReemKhan_07 Thank goodness there is no pomp and circumstance surrounding this post.

ONE word

@SusanElkinJourn Thought, just for a moment, that you meant you had played Mr Wickham (rather than (re?)viewing the play)! A reminder to us all just how much capitalisation matters.

Really liked Being Mr Wickham at Chichester this pm but omg the journey. Forgot to check trains (won't ever forget again). Engineering works between E Croydon and LGW. Coming back it was train-bus-train- tram - bus. 3.5 hours. I'd have driven if I'd known.

@Sheena2907 Do the columns of the matrix add to 0 independent of the values of x,y and z? In which case I think the matrix has to have determinant zero.

Can anyone see what's gone wrong here? I can't see why the student keeps getting det 0. Any help would be appreciated!

Merry Christmas to everyone in the National Cipher Challenge community. Hope you all have a restful day if you are celebrating, and an enjoyable day whether or not you are.

@littmath I can't help wondering how we would deal with a world in which the machines can produce and verify enormous quantities of mathematics. Who will decide what is worth reading and how will they decide that? Scarcity of resource has shaped mathematics. What will abundance do to it?

Unfortunately I don't think formal verification is (yet) a solution to the oncoming wave of AI math slop. Basic objects in many areas have not yet had their definitions formalized, let alone theorems whose English-language proofs run to 100s of pages. Hopefully this changes soon!

@howie_hua Bit harsh on Norm!

I bet no one can find a beautiful sounding name that's based off a math term. Actual mathematician names don't count. I'm talking about literal math terms. None of our terms sound good as a name.

@thetimes Looking at those boards it looks like the kids are learning quadratic reciprocity - that would be AMAZING if true!!

The top schools are preparing pupils to be curious, problem-solving and risk taking, ready for anything that comes their way — as well as academic success

@alz_zyd_ Is there any point in reading a novel now that LLMs can be trained on them? Or play chess now that Deep Mind can beat the best? How about paint a picture now that we have diffusion models? .... Who will read the papers the machines write if you don't bother learning the basics?

is there much point in majoring in math, now that the computers are better than us at math

@alex_prompter I will have to read this paper, but the list of "brittle" failures reads like a list of weaknesses shared by human scientists. Modern scientific communities have evolved a whole host of social mechanisms to moderate them and I am not sure that LLMs couldn't do the same?

This paper from Harvard and MIT quietly answers the most important AI question nobody benchmarks properly: Can LLMs actually discover science, or are they just good at talking about it? The paper is called “Evaluating Large Language Models in Scientific Discovery”, and instead of asking models trivia questions, it tests something much harder: Can models form hypotheses, design experiments, interpret results, and update beliefs like real scientists? Here’s what the authors did differently 👇 • They evaluate LLMs across the full discovery loop hypothesis → experiment → observation → revision • Tasks span biology, chemistry, and physics, not toy puzzles • Models must work with incomplete data, noisy results, and false leads • Success is measured by scientific progress, not fluency or confidence What they found is sobering. LLMs are decent at suggesting hypotheses, but brittle at everything that follows. ✓ They overfit to surface patterns ✓ They struggle to abandon bad hypotheses even when evidence contradicts them ✓ They confuse correlation for causation ✓ They hallucinate explanations when experiments fail ✓ They optimize for plausibility, not truth Most striking result: `High benchmark scores do not correlate with scientific discovery ability.` Some top models that dominate standard reasoning tests completely fail when forced to run iterative experiments and update theories. Why this matters: Real science is not one-shot reasoning. It’s feedback, failure, revision, and restraint. LLMs today: • Talk like scientists • Write like scientists • But don’t think like scientists yet The paper’s core takeaway: Scientific intelligence is not language intelligence. It requires memory, hypothesis tracking, causal reasoning, and the ability to say “I was wrong.” Until models can reliably do that, claims about “AI scientists” are mostly premature. This paper doesn’t hype AI. It defines the gap we still need to close. And that’s exactly why it’s important.

bbc.co.uk/news/articles/… @BBCnews - surely heptopus?

@bletchleypark We love our annual visit to @bletchleypark for the National Cipher Challenge prize giving and can thoroughly commend their hospitality. If you get a chance to visit, we would recommend you take it - the atmosphere is great! (Love the audio scape in the grounds)

This festive season, why not treat yourself to a afternoon tea in the Bletchley Park Mansion? Enjoy a delicious seasonal spread in the historic setting of our dining room, the perfect way to pause, relax, and make your visit even more magical. 🔗 Link in bio to book now. To make a booking using a gift voucher, please call us on 01908 272673. #FestiveTea #BletchleyPark #ChristmasAtBletchleyPark

@Millais_Maths @unisouthampton Really pleased to have you on board!

We are excited to have some teams competing in the @UniSouthampton @Cipher_Master annual cipher competition. Chapter 5 closes on Wednesday; Chapter 6 launches on Thursday. Looking forward to seeing how the leader board takes shape! #MillaisResilience #MillaisMaths

@VictoriaCoren Maybe you were thinking of wickest?

I think I must have previously misunderstood the meaning of “wicker”.

@davidbessis @skyniels I think he really meant the Lie algebra of the Pontryagin dual of the integers. (I'll get my hat.)

@skyniels You mean "a punctured S^1"

So what is the thing itself that is modelled by ℝ?