Binaries are source code outputs, they are copyrightable and patentable. Weights are not copyrightable so people can freely extract the weights and run them. If Google patents any of the novel algorithms here releasing it all freely isn't an impediment to making people license it.
Are you sure that isn't about LLMs' outputs? There I know there have been some court cases that say this, but the model itself is a work created in intricate and somewhat creative ways (I hesitate to use the word "creative" here, but would similarly hesitate to label a routine picture of the moon creative whereas pictures basically always have copyright; the bar for creativity is basically an epsilon amount above zero, afaik)
The flicker/signin redirects are a regression introduced post-Github acquisition. Really just evidence of the product being abandoned, or possibly of general operational decline at Microsoft which is affecting all their products. Definitely since the acquisition, both products have declined but Github has declined more to the point where I might prefer Azure DevOps.
Coding is a pretty small slice of the markets in play. Google's models are driving cars right now. Using coding agents doesn't give much insight into performance in the broader world; I would assume assume Google is performing better in general even if Claude or Codex is currently outperforming for coding.
> Coding is a pretty small slice of the markets in play.
I don't think that's true, mostly in that a lot of usecases are solved via coding models + a harness.
> Google's models are driving cars right now.
Yes + other models like alphafold. But those are (relatively) specialized models. Besides, the comment I was responding to was saying Google is sandbagging the market to keep it calm or something. I don't disagree that Google is doing well overall and has some clear advantages
I wasn't using Claude Code, but I told an agent to add something like this to the AGENTS.md, it did it and then a few minutes later it attempted to grant itself permission to do something and managed to delete the VM it was running on in the process. I have since adjusted the way I sandbox agents to make that less likely, but the moral of the story is clear.
Signal is designed with the assumption that data is sensitive and you should err on the side of destroying it.
Facebook is designed more as a shared scrapbook, with the assumption that data is precious and you want to share it with your community, and you should err on the side of oversharing so you don't lose any precious moments. Signal is in no way a replacement for Facebook.
> Have you used Facebook in the last 5 years? Its nothing like this at all.
I use it all the time. Yesterday I was talking to a friend, and we were reminiscing about visiting another friend's house, and we looked up some old birthday party invitations to help us remember when we had been there.
Honestly it's all there, if you use the "feeds" view in the menu it cuts out all the random influencer garbage. The search, especially the event search, is not great, but honestly I hope they don't touch it because I'm more worried about them enshittifying it further than I am about getting some creature comforts.
When Facebook was food this was how it was used and what people liked. Nobody would care what they've done to the product in the past 10 years to optimize for money over mental health
I skim a lot. I skimmed this article and appreciated the author documenting their process. I am indifferent to LLM or human writing for technical content. I suspect I skimmed most of the LLM parts, but judging writing quality was not why I read this post, I read it because I was curious about how useful the GPU is, and if I could replicate the author's work. Some carefully written prose wouldn't have helped me do that any better. The prose in this article did the job.
This is mostly how I feel about it. If anything, the weird llm jitters served almost like punctuation markers. Still, I get why it riles some people up.
I had a coworker at Amazon who always said "just do what I do, accept the meeting invite and then don't go." Linkedin tells me he's now a director at Google.
Personally I make sure meetings are a good use of my time and I complain when they are not. I also am starting to complain about AI summarizers because they frequently misrepresent what is said in meetings and they're potentially worse than nothing, although I am starting to think that they're potentially valuable if Google is trying to datamine them for info about our company meetings as a way of poisoning their datasets. But I am worried my coworkers may be thinking they are reliable.
reply