jart

25 ms·

421.
by jart 2y ago
Eventually as the tools for doing it become better they'll all want to or need to. By then, most computers will be capable of running those tools too. Which means when that happens, people will come up another way to push the limits of
422.
by jart 2y ago
I fully support using Macbooks as a thin client into a better computer. So long as it's your computer.
423.
by jart 2y ago
When I see that someone is excited about something, I believe in encouraging them. If you're looking for a more polite word to disparage people who love and are optimistic about something new, then you're overlooking what that say
424.
by jart 2y ago
I agree, and if you want to have the opportunity to build such a product, then you need a computer whose specs today are what a home server would have in four years. If you want to build the future you have to live in the future. I'm p
425.
by jart 2y ago
Using the term "bro" assumes that all AI supporters are men. This erases the fact that many women and nonbinary people are also passionate about AI technology and are contributing to its development. By using "AI bro" as
426.
by jart 2y ago
It's like asking what you're missing by not using Linux if you're using Windows.
427.
by jart 2y ago
LLaMA 3.1 405B
428.
by jart 2y ago
Or what a prokaryote would say about eukaryotes.
429.
by jart 2y ago
I hate to say it but that's like a boomer saying they never felt the need to buy a computer, because they've never wished their pen and paper goes faster. Or a UNIX greybeard saying they don't need a Mac since they don't
430.
by jart 2y ago
Every day above ground is a good day.
431.
by jart 2y ago
That's basically what companies like Mistral do. Many open source models are trained on OpenAI API request output. That's how a couple guys in Europe are able to build something nearly as good as GPT4 almost overnight and license
432.
by jart 2y ago
OSI must defend the open source trademark. Otherwise the community loses everything. The legal system in the US doesn't provide them any other options but to act.
433.
by jart 2y ago
We put a man on the moon. AI is a bubble. Nuclear power is bad. Progress comes from political showboating. Hopes and prayer let us do the impossible. I can't take any more. Hacker News is officially ruined for me the rest of today. I&#
434.
by jart 2y ago
Well this is why the rich get richer.
435.
by jart 2y ago
What do you mean by this? Why can't I just use Xcode to build my app and ask Apple for a signing key and wait for approval to publish it to their app store?
436.
by jart 2y ago
Shortcomings aren't the worst thing when you're an early adopter. I see that as an opportunity to build the #1 terminal emulator for the platform. Make myself a little money on the side. The Vision Pro could easily pay for itself.
437.
by jart 2y ago
Not when your VR headset costs as much as a Gucci handbag. The middle class won't want to join a new untested alternate reality if it's full of working class people, and the working class will always want to go where the middle cl
438.
by jart 2y ago
You ain't seen nothin. Check out the bourne shell source code from unix seventh edition. https://minnie.tuhs.org/cgi-bin/utree.pl?file=V7/usr/src/cmd... I can't believe it's not ALGOL.
439.
by jart 2y ago
I own an Apple Vision Pro and I think it's one of the most impressive pieces of technology that I almost never use, for two reasons: 1. Too much startup friction. I share the Apple Vision Pro with someone else who like me also loves it
440.
by jart 2y ago
Yeah that's what I was getting at with the tip of the iceberg thing. You can of course choose a different tip for your iceberg.
441.
by jart 2y ago
Then normal Lua by itself would probably be fastest for you. Nothing makes me happier than writing native extensions for Lua. Its C API is such a pleasure to work with. With Python and JavaScript, writing native extensions is the worst most
442.
by jart 2y ago
Use Plinko https://github.com/jart/cosmopolitan/tree/master/tool/plinko Back in the 1980s it was my greatest ambition to go on The Price Is Right and play Plinko. However all I could accomplish was
443.
by jart 2y ago
And Numpy can't come anywhere close to the performance of my C++ code. Last time I benchmarked my CPU matrix multiplication algorithm, it went 27x faster than Numpy. Mostly because Numpy only used a single core. But this was a machine
444.
by jart 2y ago
You mean the government of Prince Andrew? Yeah I think I understand now why they want the csam so badly.
445.
by jart 2y ago
Wow. I had no idea. That would explain why no one's uploaded a csam detection model to Hugging Face yet. Smartest thing to do then is probably use a model for detecting nsfw content and categorically delete the superset. Perhaps this i
446.
by jart 2y ago
If you have a csam detection model that can run locally, the vast majority of sysadmins who use it will just delete the content and ban whoever posted it. Why would they report someone to the police? If you're running a file sharing se
447.
by jart 2y ago
The government doesn't need more dragnet surveillance capabilities than it already has. Also this solution basically asks the service operator to upload illegal content to the government. So there would need to be a strong guarantee th
448.
by jart 2y ago
Are you saying someone will use it to create a CSAM generator? It'd be like turning smoke detectors into a nuclear bomb. If someone that smart wants this, then there are easier ways for them to do it. Analyzing the detector could let
449.
by jart 2y ago
So they publish an updated model every three months that works better.
450.
by jart 2y ago
If governments and big tech want to help, they should upload one of their CSAM detection models to Hugging Face, so system administrators can just block it. Ideally I should be able to run a command `iscsam 123.jpg` and it prints a number l

More ›