Christine Lemmer-Webber

One of the top threads on the ClaudeAI subreddit right now: "I am done with this shit" https://www.reddit.com/r/ClaudeAI/comments/1wm5c21/i_am_done_with_this_shit/

Curious to see just how much of the thread on a quite pro-genAI subreddit is people commiserating on the exhaustion of their work turning primarily into tedium and poor engineering.

Bonus in that the Claude AI summary bot in the thread says:

The Rise of the Meat Proxy: You're not an engineer anymore, you're an "organic vessel designed to press enter." Or, as one user so poetically put it, you're just a condom between Claude and your production server getting fucked. Your only real job is to be the warm body they can blame when the AI slop inevitably breaks.

It's pretty simple. You're hearing "We're so scared of the AIs, we need to reign them in" from the very US companies building and selling them for two reasons:

• They're about to IPO, and it's amazing marketing for them
• They want a regulatory moat, since they're deeply worried about the open weight LLMs. They want everyone dependent on this tech and to be the only ones legally permitted to dispense it.

It appears that Tristan Buckmaster raising alarms that AI corporations may have snooped on their research so the companies could throw compute at developing proofs for, and gain credit for, advancements in the Navier-Stokes equations. OpenAI then threw a massive amount of compute at the problem, and seemed to try to coerce Tristan into getting credit for it: https://cims.nyu.edu/~tristanb/statement.pdf

I said that if OpenAI released its result in the way proposed I would go public with what happened. The reply was, “Why would you ruin your career?” I replied that I am an academic, and asked why he thought going public would ruin my career. The reply was, “If you don’t want me to be nice, then I don’t have to be nice.”

Oh, OpenAI has responded now:

On Tuesday, September 1, we heard rumors that two Millennium Prize problems had been resolved. Inspired by these rumors and by the step change in performance of our internal model, we launched an effort to evaluate it on all open Millennium Prize problems and a few other high-impact problems.

https://openai.com/index/navier-stokes-solution/

"heard rumors" 🤔

P.S. Anthropic’s C compiler and Cursor’s FastRender web browser haven’t had any commits for months.

I feel like some people may not have met Christine, and you should know that she never jokes about technical matters and is utterly serious at all times. https://codeberg.org/guix/guix-consensus-documents/pulls/14#issuecomment-20224132

Yes that's right thank you @bremner@mathstodon.xyz