somebody looked at Claude Code's binaries and Anthropic is testing out their own app platform called antspace. Not sure why people are shocked, they've been cloning features of their API customers and adding them to their core products since day 1. Makes sense they will take user data and do it for Claude Code by copying features or buying up what developers are using so they can lock people into a stack. These are the same people that trained on every scrap of data they could get their hands on and now complain about distilling models from their output
Ironically this type of stuff really makes me doubt their AGI claims, why would they bother with this stuff if they were confident of having AGI within the next few years? They would be focused on replacing entire industries and not even make their models available at any price. Why bother with a PaaS if you think you are going to replace the entire software industry with AGI?
cell tower dump and general geofence warrants to Google/Apple were how many Jan 6th protesters were found and charged. Courts threw out all their arguments about this very issue. This is standard practice and was celebrated as cops being smart
here's a Washington Post article lamenting that Google was cutting back on how long they hold location data and how hundreds of people wouldn't have been prosecuted without it- https://archive.ph/r7afb
you'd think they could use AI to interpret the best model for your use case so you don't even have to think about it. Run the first few API calls in parallel, grade the result, and then send the rest to whatever works best
Gwern talked about it when GPT 4.5 came out, a ton of breakthroughs with image and text AI came from anons trying to optimize models for their waifus. That's basically the origin of chat models in general - https://www.reddit.com/r/OpenAI/comments/1izpgct/comment/mf5...
those endowments, especially for the Ivy League schools, aren't liquid at all. They'd take a massive haircut if they had to start pulling funds from it
no, but there would be much more push back against this type of action if Harvard and other universities didn't alienate a large chunk of the population. Why should the taxpayers fund places that openly admit to decades of racial discrimination in admissions
the institutions have already failed their intended purpose, as shown by the research fraud. Propping them up with tax dollars because of nostalgia over the name brand is pointless
the judge you are quoting literally worked in FDR's admin when they were deporting millions of Mexicans, regardless of whether they were born in the US. They didn't get due process
these types of moves wouldn't be possible in the first place if these institutions hadn't spent decades burning their own credibility. They even mention Alzheimer's research in this post, something that has literally wasted billions of taxpayer dollars due to an academic cartel shutting down anybody trying to expose the fact that they were completely wrong about amyloid plaques
just cutting government spending would automatically put the US into recession, deficit spending is the only thing that has kept things going. Huge amount of job growth is either direct government hiring or indirect through businesses hiring off revenue from government contracts
hyperscalers in shambles, no clue why they even released this other than the fact they didn't want to admit they wasted an absurd amount of money for no reason
saw someone point out how they could have framed it as a way to reward good workers, but instead they showed how they can go after people underperforming. Just bad framing and shows how out of touch they are
not sure why people are surprised, it's been known a long time that RLHF essentially lobotomizes LLMs by training them to give answers the base model wouldn't give. Deepseek is better because they didn't gimp their own model
comparative advantage is disproven at this point, because economists are failed math majors who don't account for the innovation flywheel effect and general network effects. Silicon Valley software industry happened because of the proximity to the existing hardware industry there. Same thing for other forms of manufacturing and the US gave it away in the name of short term profits
Gen Z only voted democrat by a 4% margin which is an absolute death sentence for the democrat party considering people only vote further right as they get older. For context Obama won that demographic by 20% in 2012, and won it by even more in 2008. If those numbers hold democrats have no path to winning national elections
https://x.com/AprilNEA/status/2034209430158619084
Ironically this type of stuff really makes me doubt their AGI claims, why would they bother with this stuff if they were confident of having AGI within the next few years? They would be focused on replacing entire industries and not even make their models available at any price. Why bother with a PaaS if you think you are going to replace the entire software industry with AGI?