"All models are wrong" - not mine. Mine are right. Perfect, some even say. I make them purely from the data of the thing they're modelling, making them useless, so that's another thing wrong in your saying.

Roman gladiator devouring the flesh of his combatant, 50 BC

Some businesses just don't understand friction. Searching for storage units in my town, 8 out of 9 needed my contact so they could call me and demand more info to give me a quote. 1 put the quote on the website, where I could book directly. Guess who got my business.

The idea that academic works should be immediately useful or even read misses the point. Sometimes I read papers written decades years ago and find useful things. That’s what “adding to the body of knowledge” is. You cannot judge a work by its contemporary popularity.

People complain about twitter since The Change but I've found it useful for A) finding cool papers as they come out, and B) arguing with people about physics, which are like my 2 main drivers in life right now

Hello everyone. A friend told me that I shouldn't post this message because it made me and other PhD students look bad. But I actually think it's important to show how PhD students (especially foreign ones) have to deal not only with research-related difficulties, but also with

AI being a superhuman lit reviewer for mathematical results checks out as a legit superhuman purpose of it. Other people claiming it's *proving* new results without understanding the distinction checks out as a meat-filled amoeba-descendant skinbag level of intelligence also.

I wonder if maths is uniquely positioned to be prone to AI hype. New (actual) results are so abstracted away from most people's day-to-day lives that they can cheer on misdirected AI praise while not caring about the actual content.

So great to see a return to the old art forms. Heisting through a window to nab Napoleon’s jewels. I hope they wore cat burglar costumes. jrnl.ie/6848478

I get that some people are upset by this, but these items just sit in a museum unused by anyone. Anything that adds an air of intrigue to them - like this theft, damage and discovery on the street definitely does - only makes them more valuable as museum pieces.

Paris did this weird thing where they restricted driving and the air is noticeably less smoggy and clear skies seem clear, but you can still sit at a street side bistro and get bombarded with traffic sounds and watch the deadlock not go by.

AI is a good litmus test for how well one models the function of education. If you want to “get into” AI, should you do a grad degree in it or in mathematics or something instead? Obviously the latter, as whatever non-fundamentals you learn in the former will be outdated soon.

The word for “quark” comes fro James Joyce’s “Finnegan’s Wake” - “three quarks for muster Mark”

Was travelling through Paris y’day and saw a really shiny gemstone half submerged in a puddle. Was about to pick it up when I remembered that real gemstones aren’t just lying around on dirty city streets! Always nice when logical deduction saves you from looking like a fool :)

Year 2031. I, a young PhD student, submit my thesis to the AGI. It whirs for 1.3 seconds, then stops. "W-what do you think?" |Yeah. Seems fine| "Is that- do I pass?" |I've incorporate the 3.2 useful sentences into the body of knowledge. Please buy more tokens for feedback|

Lot of talk about the Irish Presidential Race and the weirdness (in a bad way) of it this year. What will only hit afterwards is that Michael D will no longer represent us, and I'll feel some genuine sadness about that. He was my first ever vote!

Everyone mentions how AI text has lots of em-dashes and “not just _ but _”, but has anyone figured out why AI video is like 60% fisheye lens?

Can’t wait for AI to satisfy my hunger to watch human-esque facsimiles doing not-quite-human things, because I, medically, am a psychopath.

The dictionary admits its cardinality is at least N, meaning they're shorting us on their printed books. If I see a π in there then we really ought to complain.