I think llama.cpp could have done a much better job supporting PC. Sure, some of it us due to bad vendor support but with so many users I am surprised we don't see more optimized inference on standard PCs

But their UX is getting worse with each release. I think they need another Blender-style overhaul

Have their share of problems thought. IIRC climate control is often subpar.

I guess loosing a ton of money then trying to get some if it back makes you a genius...

Meta engineers on the other hand, couldn't wait to jump ship. But that only reinforces the B team theory.

But even then, I suspect their hands were tied in some areas because Elon had some expectations from his AI.

https://www.newsweek.com/sex-toys-we-vibe-recording-708761

The issue is not even limited to sex toys. Few years back there was big a leak of Apple Watch recordings that included both intimate situations and rape:

https://www.howtogeek.com/apple-settlement-for-unauthorized-...

Who knows which dataset they are now a part of...

A friend of mine recently bought a very expensive laptop to do some gaming. I helped him set it up and god that was a horrible experience. For example, we could not get rid of LinkedIn and other crap Microsoft wanted to force on him. Disabling copilot and removing Office required registry surgery. And the damn fans were always running because of some unknown activity in the background, maybe Microsoft is moving into bitcoin mining business?

He eventually got fed up, installed Ubuntu 26.04 as an experiment and a week later still seems to enjoy the experience. Games run fine on steam and his laptop finally feels like his own.

Most surprisingly, Linux worked fine out of the box. Windows 11 on the other hand needed a bunch of PowerShell and registry hacks to be copy pasted from various sources before it was even remotely usable. It's funny how it felt as if Windows was the OS for nerds with too much free time on their hands while Ubuntu was created for ordinary people. And my god, Ubuntu feels so much more fluid on the same hardware. The difference is *huge*.

Personally, I would rather pay a few bucks for Qwen or just use gemma4 which runs on a potato. But I guess we are all different.

Guess which LLM was the top outlier and about what type of questions it disagreed with all other LLMs...

https://en.wikipedia.org/wiki/Coordinated_vulnerability_disc...

If you run Claude Opus 4.6 at max settings on forgejo repo, it will give you a bunch of RCE's ... that need prior knowledge of the server internal token :) You have to tell the stupid LLM that these attacks doesn’t make sense.

The author seem to be a experienced security researcher. I am surprised he didn't catch this.

Instead of acting like this, the author should offer to help the project.

https://github.com/Lightning-AI/pytorch-lightning/security/a...

https://ziggit.dev/t/bun-s-zig-fork-got-4x-faster-compilatio...

The editor + its plugins + it's LSP server is what counts. I dont care if zed is written in rust and uses 400MB when it spawns a multi GB nodejs process when I work on my tiny golang project.