schmorptron

6 ms·

1.
by schmorptron 6d ago
They seem to be getting pretty far with only vision, to be fair. Same with openpilot. Still, I cant help but agree. If you have access to beyond-human sensors and are able to do sensor fusion, then please do it. What's the point other
2.
by schmorptron 6d ago
"here at waymo we hire a list of fall guys who will go to jail if one of our cars hits someone. its a great job, just like being oncall"
3.
by schmorptron 8d ago
Awesome, thanks!
4.
by schmorptron 8d ago
It's kind of insane how having this tech at this speed 5 years ago would have probably still been seen as insanely useful and revolutionary. If LLMs were more capable but dramatically slower, I wonder how it would impact how we use it?
5.
The Ripgrep of AI Context(github.com)
2 points by schmorptron 8d ago|0 comments
6.
by schmorptron 8d ago
I'm not the author, but I would like! Is it feasible to run on the same hardware as the 27b model itself?
7.
by schmorptron 14d ago
A reminder that google is the only major lab without a meaningful opt-out of training on your data. The only way to opt out is to disable message history entirely, which seems like a darkest of dark patterns to get users to leave "opt
8.
by schmorptron 16d ago
Within the general theme of token subsidies slowing down, this seems like the logical step. In my mind the timeline goes something like this product wise: - Pepole figure out that having a general purpose not-just-coding agent actually wo
9.
by schmorptron 20d ago
I'd prefer that too. Preferrably with permission popups for requesting folder access outside of it. Shared libraries / whatever flatpak calls them could live at a central symlinked location?
10.
by schmorptron 23d ago
Really cool project. I think it's nice that there's so many small, experimental distros out that are just trying stuff and seeing what works and what doesn't. If we (maybe through where freedestop is moving) can figure out a
11.
by schmorptron 29d ago
Agreed. I've run into this way too many times, and now with a super aggressive oom killer config it still slows down until vscode or firefox just get closed. I wish it was similar to how it seems to be on MacOS from what I see on frien
12.
by schmorptron 1mo ago
This is what the author showed off on Twitter at least, with Desktop chromium I think.
13.
Tess's Android Wayland Compositor(github.com)
86 points by schmorptron 1mo ago|17 comments
14.
by schmorptron 1mo ago
The past DeepSeek models and now these new checkpoints score very badly on the ArtificialAnalysis AA-Omniscience and hallucination rate benchmarks. I wonder where that's from? Maybe they're overindexing on coding even more than ot
15.
by schmorptron 2mo ago
You make a good point about cultural things that would make one excited to learn a language. But I really struggle to imagine what could be done to fix that? Injecting more money into the film industry would just lead it to the same people
16.
by schmorptron 2mo ago
What a nice and heartwarming read! I'm german, and really hope we can pivot away from the rise of the far right. How that can be done, I don't know. I just hope we can do so before we start scaring away people like him who would o
17.
by schmorptron 2mo ago
App tells customer to review it. Customer reviews the app honestly (the nag is annoying and deserves 1 star). Seems like the nag screen worked? :) (Although in seriousness, I get this point they make further down in the article: "If we
18.
by schmorptron 2mo ago
it has to be so bursty for realtime usecases like chat, which is what most people are using it for today. of course, once (if) stuff like software dark factories start working out for the average person, then you'll be able to make ful
19.
by schmorptron 2mo ago
LLM inference unfortunately also seems to be a task that's poorly formed for moderate consumer hardware,as a single user. For a single user use case, the load is bursty but requires the weights to be in memory already. So a multi user
20.
by schmorptron 2mo ago
It's impossible to keep ads "clearly labelled" and seperate long term. Doesn't work. the incentive structures just don't work. serving answers becomes a cost center, and ads are a profit centers. so even though it m
21.
by schmorptron 2mo ago
Yeah, this has been my progression as well. Maybe next it'll be just using plain pi when you figured out exactly what you want from omp and what you don't
22.
by schmorptron 2mo ago
Are thinking models only the reasonable tradeoff vs using much larger non thinking ones because the cost of output tokens is below that of input tokens?
23.
by schmorptron 2mo ago
That's a more than 2x jump in parameter count. I know it's not a measure of quality by itself, but it will be interesting how it "scales". Bust it looks like they're gonna be competing with the big boys now, pricing
24.
by schmorptron 2mo ago
yeah, but that's due to enterprise commitments that MS won't train on the user interactions
25.
by schmorptron 2mo ago
I don't think that's it for console manufacturers. They make the majority of their money on game sales, so they want the console itself to be used for as long as possible.
26.
by schmorptron 3mo ago
We're moving towards total surveillance slowly but surely. Age verification. Chat control. To an extent also the digital euro. It all seems hopeless, they're pushing this through despite what semblance of a democratic process we h
27.
by schmorptron 3mo ago
i love that word, and now it's genuinely (hehe) ruined. thanks, claude
28.
by schmorptron 3mo ago
I'm giving them the benefit of the doubt and interpreting it in a charitable way because they sound earnest about it, this is incredibly ambitious and cool-sounding, and I wish them all the best. It's something that's some so
29.
by schmorptron 3mo ago
you can build the datacenter right next to the tank and use the now-warm cooling water to pump into the tanks!
30.
by schmorptron 3mo ago
Cursor's composer models are finetuned kimi

More ›