I use it as a hosting target for automated deployment tooling we wrote. Tools were originally Digitalocean-only and I wondered if LLMs could add support for a second provider, so asked Claude to add Vultr which it did very nicely. But other than run the automated tests (which create VMs, DNS entries, check validity and tear down) and pay the monthly bill, we haven't yet used them for production deployments.
Linux is only widely used/successful today due to commercial vendors who consumed it as way to reduce their input costs. Arguably the same incentives exist for AI in which case the way forward would be through some sort of consortium of companies that use AI themselves funding the creation of models. Obviously when this scheme is extended to the limit, you get governments funding model creation, much like they fund building of roads, railways, ports and atomic weapons.
These things sound plausible, but have they actually been demonstrated? Wouldn't anyone who succeeded in making such a small but useful LLM be raking in the money now?
Cursor's composer 2.5 is a perfect example. It's right on the heels of the frontier (for coding only) for an order of magnitude cheaper. As much as I've shit on Cursor in the past, I do think the company is well positioned to pick up people getting sticker shock on Anthropic tokens, if they can get their marketing down.
It is, but the US labs have been pushing parameters heavily. There was a pullback from big models after GPT4.5 in particular, but with a shift towards emphasis on post training and the good results Google got with scaling Gemini 3, all the labs started to push scaling again, which is the reason the frontier is getting more expensive. So that 1T isn't as big as it sounds, the American frontier is probably sitting at 3-5T at least.
I think Elon pretended to notice it. Instead of taking his time and doing it right, he was performative in his own actions causing unnecessary problems. In doing so got rid of the important pieces as well as the bloat.
It is actually shocking that twitter is still standing after his severe headcount cuts. I have not yet read an analysis of that. How was the system able to keep going with almost no downtime after such severe layoffs?
Why does everyone focus on this aspect? Why is this surprising? Do people think that 100% or even 20% of Twitter employees were SREs? Do you think that most large applications are kept alive by constant manual toil from SREs? (ok, ok some are - but still!)
What's funny is that Twitter SRE used to be horrible and the app probably would have collapsed entirely (rather than the little bit that it did) without hundreds of manual operators, but in the few years leading up to the "acquisition," massively improved to the point that they literally automated themselves out of a job.
Anyway, Twitter had thousands of engineers, salespeople, support people, and so on. They were working on tens of new products in an attempt to find more revenue (everything from clones of every single social media app you can imagine to becoming a sports TV network), and on the other side (Goldbird), selling and supporting ads, the thing that made Twitter money.
The metric to look at isn't uptime, it makes no sense that people keep parading this metric. The metrics are revenue and revenue growth and surprise! by most available metrics, the Elon strategy torpedoed those.
Twitter was, like almost every "web" company in ~2020, a very "fat" company because they were re-investing free ZIRP money in future growth investment. Elon turned it into a KTLO operation, and didn't even manage to succeed at the standard PE style "fat" company slim-down (where you chop growth initiatives and keep the revenue, like everyone else is doing now), because he also chopped the revenue side.
It's obviously hard to say how accurate these numbers[1] are, but it looks like Twitter has doubled their workforce from its lowest following their mass layoffs. It might be stable again now because they hired the workforce required to actually keep it running.
Is it really that shocking to you? Twitter is a very narrow company compared to Microsoft, Meta, Google, and Apple. The system was already up and running, they probably kept the employees that built or knew the system deeply and fired the others. Apparently it had around 7,500 employees at its peak. To me that seems excessive for something like Twitter.
reply