If the proof of RH is confirmed, what dependent conjectures becomes proven? Someone compiled a list of consequences. Factoring large composites is still difficult, but now there's a upper bound, and perhaps faster more capable computers will be able to factor them. So the world is safe, for now.
––– 凄い –––
Someone thinks that Alpöge's proof is invalid. But not to worry – other mathematicians have attested that the proof is indeed valid.
Another person has developed a human readable (non-LEAN) version of Alpöge's RH proof. It's still very abstruse, of course.
Remember that independent researcher who claimed to have a proof of RH earlier this year? Well now he's suspecting that Claude plagiarized the proof he posted online. His proof, which relies on polyhedrons, geometry, the Golden ratio and nuclear physics, doesn't even resemble the approach that Alpöge took.
Possibly because of the growing interest LEAN, someone put together a tutorial for beginners.
Oops. Did Anthropic and its bankers get a key call wrong after IPO? There's a mechanism called a greenshoe, which helps protect the share price of a company in the early stages of an IPO. If the share price takes off, the underwriting bank is guaranteed the right to buy some extra shares at the IPO price. If the share price tanks, then the underwriting bank has 30 days to buy a bunch of shares on the market to cover its initial short position, using money that it gained by overselling shares at the start of the IPO. The buy signal stabilizes the price drop and prevents embarrassment. But apparently Morgan Stanley exercised the greenshoe the day after the IPO. Now it's stuck with overpriced shares. Which means that if the stock price continues to plunge, it has to use its own cash to send out that buy signal, which means taking on risk. Smart people are supposed to be able to avoid that.
––– 良くない–––
iLoveNotch is an open source utility that can turn your MacBook notch into a launch pad of source for all kinds of functionality. Could be useful.
––– 凄い –––
Maple AI looks like another private cloud compute service. I can't tell if it's like Together AI with ZDR, or like Ollama, where you can run open-weight models entirely on your own computer. Perhaps both? Long-term memory could be very useful, so I don't have to build it myself.
––– 凄い –––
Moderate coffee intake linked to lower risk of recurrent VTE, particularly after provoked index event. Moderate intake is 1 to 2 cups/day and the hazard ratio is 0.5. But if you drink more, like >6 cups/day, the protection decreases, with a hazard ratio of only 0.71, which is still something. It's only for provoked clots. I'm always suspicious of studies like this when there are exceptions. It makes me think the observation was just random and a chance finding. Why is there no dose-response relationship? Why only provoked clots?
––– 凄い –––
Hope springs eternal. Oregon’s job market may be looking up, finally. I had to laugh at this. Just a little turn at the end, and the Oregonian thinks that things are getting better. I don't see a trend. Do you?
––– 良くない–––
Portland city workers strike drags on as another union threatens walkout. This is why nothing gets done in Portland. And what do you think will happen with the money the city gets from fees and taxes on CEOs?
––– 良くない–––