accidentalflyer ๐Ÿ‡จ๐Ÿ‡ฆ๐Ÿ‡น๐Ÿ‡ผ๐Ÿ‡บ๐Ÿ‡ธ๐Ÿ‡บ๐Ÿ‡ฆ

@accidentalflyer.bsky.social

A poor wayfaring hedge knight in feudal Silicon Valley who watched C-beams glitter in the dark. The truth is not kind And you've said, neither am I PhD Aero. Opinion Zero. Also b-school, and jester of pain. Small time amateur pilot.

๐Ÿšจ๐Ÿšจ๐Ÿšจ Seems like another Mathematician comes forward with concerns about OpenAI stealing unpublished math research and then publishing it as AI solve under big publicity. Andreas Thom and Gรกbor Kun worked on Gromovโ€™s soficity conjecture. (See his thread with alt-text) mstdn.science/@andreasthom...

1/3 I want to record a number of points that I realized after reading up on the controversy. In fact it recalls an exchange I had with OpenAI after its non-sofic-group announcement. It exposes a serious, unanswered question about whether researchers can trust OpenAI with unpublished mathematics.

I wrote in an email to Mark Sellke and Sebastien Bubeck shortly after the surprising finding of a non-sofic group using the methods of Kun and myself:
โ€œAnother point is that I and a colleague in Dresden were discussing the expander matching problem and various extensions of the work with Gabor Kun actively over the last months with ChatGPT, so that we are of course curious if that was part of the training data or accessible to the reasoning process. There is a certain (frankly unacceptable) lack of transparency here; and I fear it will damage the communal process of math more than the new AI-generated results will benefit the subject.โ€

Mark Sellkeโ€™s complete answer was: โ€œRegarding your conversations with ChatGPT: that did not happen.โ€

I had explicitly asked about two different things: (1) whether our conversations entered training data, and (2) whether they were accessible to the solving process. The categorical answer now looks as though it addressed only direct access under (2). No such qualification, explanation, or evidence was given. I take this as dishonesty to say the least.
OpenAI now says in the Buckmaster-Alpรถge case that no specific user data was accessed, but adds that it โ€œcannot rule out that de-identified data derived from their usage of our products helped improve our models.โ€
Philipp Markolin@philippmarkolin.eurosky.social ยท yesterday

The good old: "If you go public, just be careful not to destroy your career and btw, we will be done playing nice" Original statement from Tristan (NYU Mathematician): cims.nyu.edu/~tristanb/st...

Zelensky says his visit to Canada will include talks with Prime Minister Mark Carney focused on strengthening Ukraine's air defense and preparing for winter. He says planned meetings and agreements will reinforce the strategic relationship between Ukraine and Canada for years ahead. #Ukraine

Silicon Valley is very much not just founders and investors. When it comes to innovation those are often the least important people. Even just focus on the tech industry, results come from the employees with hands-on skills who do the actual, "hands dirty" work.

Next door in Silicon Valley@nextdoorsv.bsky.social ยท 13h ago

jfc. Silicon Valley is more than just people who either do 2 and 20 fee structures or are at Y Fake It To You Make It Combinator. Give me a fucking break.

Remember when Hillary Clinton told a story about landing in Bosnia "under sniper fire" when actually it was just a war zone but there was no record of direct fire and the media made it into a multiple-week scandal? Only Republicans are allowed to flagrantly & repeatedly lie.

Trumpโ€™s top six falsehoods about 9/11 as 25th anniversary nears

US president often brings up al-Qaida attack that struck his home town New York City almost 25 years ago

theguardian.com

BREAKING: Jimmy Kimmel says his interview with Texas Senate candidate James Talarico will NOT air on ABC due to threats from Trumpโ€™s FCC. The interview will instead be posted to YouTube.