Europe Pub

cross-posted from: https://endlesstalk.org/post/91955305

Archive: https://archive.ph/j7eiz

Coders spent more time prompting and reviewing AI generations than they saved on coding. On the surface, METR's results seem to contradict other benchmarks and experiments that demonstrate increases in coding efficiency when AI tools are used. But those often also measure productivity in terms of total lines of code or the number of discrete tasks/code commits/pull requests completed, all of which can be poor proxies for actual coding efficiency. These factors lead the researchers to conclude that current AI coding tools may be particularly ill-suited to "settings with very high quality standards, or with many implicit requirements (e.g., relating to documentation, testing coverage, or linting/formatting) that take humans substantial time to learn." While those factors may not apply in "many realistic, economically relevant settings" involving simpler code bases, they could limit the impact of AI tools in this study and similar real-world situations.

Coders spent more time prompting and reviewing AI generations than they saved on coding. On the surface, METR's results seem to contradict other benchmarks and experiments that demonstrate increases in coding efficiency when AI tools are used. But those often also measure productivity in terms of total lines of code or the number of discrete tasks/code commits/pull requests completed, all of which can be poor proxies for actual coding efficiency. These factors lead the researchers to conclude that current AI coding tools may be particularly ill-suited to "settings with very high quality standards, or with many implicit requirements (e.g., relating to documentation, testing coverage, or linting/formatting) that take humans substantial time to learn." While those factors may not apply in "many realistic, economically relevant settings" involving simpler code bases, they could limit the impact of AI tools in this study and similar real-world situations.

by Caitlin Johnstone
Jul 18, 2025

It’s 2025. Israeli soldiers are telling the Israeli press that they’re being ordered to massacre starving civilians trying to obtain food from aid centers. Countless doctors have been telling the world that Israeli snipers are routinely, deliberately shooting children in the head and chest throughout the Gaza Strip. Amnesty International, Human Rights Watch, and all the leading genocide experts and human rights authorities are saying that a genocide is being perpetrated in Gaza. The New York fucking Times just published an op-ed by a Zionist genocide scholar who’s finally admitting that it’s a genocide.

There’s no way to deny what this is anymore. If you still support Israel in the year 2025, it’s not because you don’t believe Israel is committing horrific atrocities. It’s because you believe those horrific atrocities are good, and you want to see more of them.

cross-posted from: https://lemmy.eco.br/post/15004489

bluesky source

artist's patreon

cross-posted from: https://leminal.space/post/23122926

Gringos out here wilin'

That would've been more successful if it released 5 or 10 years or even more later.

https://tidal.com/browse/track/668138?u

bluesky source

…with carrots, mushrooms and onions and marinated in sake before cooking. Served over plain rice. Delicious even though the mushrooms got stuck under the skin.

After a ProPublica investigation revealed how Microsoft’s “digital escort” tech support service could expose sensitive government data to cyberattacks, the company says China-based engineers will no longer provide assistance on DOD cloud services.