Oh-oh. There's a rumor that OpenAI may have "trained Astra on conversations in which he [Andreas Thom] and Gábor Kun were working on Gromov’s soficity conjecture". If this is true, people REALLY have to keep their research away from OpenAI's models. Lest your research career vanish in an instant.
Here's another post about the Navier-Stokes situation. What an experience for an academic.
––– 良くない–––
There's a lot of discussion about the resignation of the a guy named Coxon, from Anthropic. But here's another take. Was it just a smokescreen? Look at the list of Dem senators who are now saying that AI regulation is justified.
And here's Meta, launching Muse, which will, among other things, send emails and money from your account automatically. Yeah, who needs this crap?
––– 良くない–––
The Four-Color Theorem Gets a Rare New Proof. The original proof by Appel and Haken that was done with computers was kludgy and unsatisfying. Now there's a new one which is less kludgy. It still uses computers and checks over a list of graph configurations. I guess for problems like this, you just have to put up with a little bit of inelegance.
––– 凄い –––
Caltech is planning a math hackathon, or Mathathon, but a bunch of Caltech and non-Caltech mathematicians aren't happy about it. They don't like it that participants are being encouraged to use AI to solve problems, as they are given AI tokens. Plus, it may scoop out existing mathematicians who may see their research significantly deflated, and their ideas scooped. Plus, who is going to verify the results? Professional mathematicians, of course, who likely will be unpaid and uncredited. Also, the results won't be posted on preprint servers, so any insights will be hidden. But how do you verify AI-generated results? So far, the contest is still on.
This is similar to the kind of angst that some physicians are dealing with. When you're develop the AI and own the tools, you're happy. But for users, it's just a brave new world.
––– 良くない–––
Remember in 2000 when the world was looking for Finland as the answer to improving math and science education? Finland had the answers, it seemed. Well, it turned out that it was a fluke, and Finland's PISA scores have been declining ever since. And look at the latest PISA scores, just released. As always, the top performers are in Asia. But does the West look to Asia for tips? I don't see that. I find it ironic that you never see Chinese, Japanese or Korean persons heading education departments in America? Wouldn't it make sense if they did? Our academic institutions seem to exclude them instead. Very strange.
––– 良くない–––
Anthropic reveals fourth likely crime committed by its AI. We're getting inured to this, aren't we?
––– 良くない–––
There's a guy named Elizier Yudnovsky who famously made doomsday predictions about AI. Turns out he was wrong almost all of the time. He's like the Jim Cramer of AI. 🤣
––– 凄い –––
I never heard of the 7 Vibe 3 lifestyle. Apparently that's how modern entrepreneurs see their work-life balance. Whatever.
––– 良くない–––
iOS 27 RC Adds Long-Rumored Camera App Controls Customization. There's some of this already. I don't like the existing UI, so I hope this improves.
The new Apple Watches will have ears. Mobile phones changed the world when they had cameras built in. Now they will have audio recorders built-in. Not sure I like that. Apple says not to worry, but I can't be sure.
And Apple is introducing a feature where you can store a reference image on their servers, to convince others you have the real photo, and not AI slop.
––– 凄い –––
NASA’s New Experimental Jet Promises Supersonic Flight Without the Sonic Booms. That's great, but look at the needle nose on that thing. You could take an eye out!
––– 凄い –––
Not to be outdone with features, Google's Pixel Watch will detect if you stop breathing, and notify people if that happens. Apple has something like that already if you fall down.
––– 凄い –––