ritchie

@ritchie.bsky.social

Recovering academic. I make tools for facial animation and game narrative at JALI. (@jaliresearch.bsky.social)

I feel like no one who follows me on Bluesky cares about this but I’m very intrigued about the issues with formalization / reliance on prior proofs in OpenAI’s new non-sofic group proof. Basically in their write up they said they were relying on a prior proof that then people found to have errors?

Fluke Seawalker has stray ketchup packets, dumpster noodles & partially lead/cholera-free water for dinner. Call of Colonialism's FBI warnings say that 'Winners Don't Use Drugs -- except non-generic opioids'. Will Fluke survive the toxins & toxic misinfo? Find out in Neofeud 2! #indiegame #art #noai

final analysis: "more there than I remembered. very rough around the edges... and the end... and the middle... intriguing. it's amazing to me Boatmurdered even got written, given how rough that is"

Between the news about Miller and Platner over the last few months, I'm going to share my deeply informed position that Marines should probably stay out of politics. There's arguably a reason the Marine Corps is over 250 years old and we've never had a Marine as president.

There’s something especially perverse about computing being one of the most autodidact-friendly intellectual fields ever created, while its professional culture keeps reinventing entrance exams for an imaginary guild.

fucking fightin' for my life over here scrapping together open source datasets and I've gotten pretty good at it but then you see articles like 'with the cooperation of the prime minister's office, we randomized which cities received the subsidy to ensure causality' like ok guess I'll just die then

ai haters: why are you using ai ai boosters: why are you insulting ai people who actually know ball: all code is horrible, none should ever be made in any way

this is getting some traction because someone reposted it - looking at it again, I want to point out it doesn't mention if the trends for men are statistically significant, and so I don't think they're conclusively shown by the paper (definitely they're much much weaker trends than the main results)

ritchie@ritchie.bsky.social · last mo.

And from the appendix: Men do not increase their partnership rates, but if they’re in relationships they leave their partners instead!