andai

13 ms·

361.
by andai 2mo ago
"Computer. Search result. Relevant."
362.
by andai 2mo ago
AGI ≈ Ralph × Infinite persistence Just like real life!
363.
by andai 2mo ago
Thanks for the effort you put into this. I wonder if it might drive the point even further if the graph scales were linear? Or maybe the progress has been so great that this would make the graphs unreadable? https://xkcd.com&#x2F
364.
by andai 2mo ago
> Reading everything becomes the default. At a cent per document, a model can read every paper I love how in our day "reading everything" means "the computer reads it for me". I expect soon the computer will be able t
365.
by andai 2mo ago
Yeah they're not great at searching, but they make up for it by hammering 2,000 websites in parallel for ten minutes straight. Gets the job done!
366.
by andai 2mo ago
Has anyone tried making a p2p search engine? Naively it seems like one of the p2p networks from the late 90s or early 2000s should be able to handle it. Searches propagate between nodes, nodes specialize on specific sections of the index, w
367.
by andai 2mo ago
I saw a 1B model yesterday that was fine tuned on Fable output. I found that hilarious, but it did actually make all the scores go up. (Actually talking to it, it was about as coherent as you'd expect, i.e. 3/10) The floor for &qu
368.
by andai 2mo ago
A year ago I had an aha moment, when I realized that for my purposes, Gemini Flash was not only 9x cheaper, but 3x faster than Gemini Pro, while producing identical output. Who's the best model now! For a lot of tasks, even small model
369.
by andai 2mo ago
Wait til the robots get on Cerebras, it'll set your pants on fire.
370.
by andai 2mo ago
I had to grope a little, it was in the External Links section! https://web.archive.org/web/20200717054958if_/https://www.is...
371.
by andai 2mo ago
We plateaued about 10 years ago right?
372.
by andai 2mo ago
What does native mean in this context?
373.
by andai 2mo ago
Yeah, I heard most kids these days can't read analog clocks either. I can't actually remember where I learned to read a clock, it might have actually been in school. I guess that means they don't teach it anymore. (Everyone&#
374.
by andai 2mo ago
They process the original image file with Python on the local device. (And I've seen the web chats do this with their "computer use" features too.) The really wild one is even blind models will do this and they'll try to
375.
by andai 2mo ago
<html> <head></head> <body></body> </html> That's funny I thought it would be an empty response, but they explicitly hand-coded it like that. Also -- offtopic but I think
376.
by andai 2mo ago
Is there an expectation that OR customers will jump ship due to something about Stripe, or about acquisitions in general?
377.
by andai 2mo ago
I had a LLM based prototype called "cleanroom" which would convert a program into a spec and then back into a program. The results were disgusting: the spec would encode all sorts of irrelevant implementation details, and then the
378.
by andai 2mo ago
> There’s no reliable record of human intent. Prompts are discarded, and the code may or may not have been generated by AI. We’ve lost the central authority that expresses what the human wants out of the machine, and I think it’s importa
379.
by andai 2mo ago
https://youtube.com/watch?v=wgPf24_OQZY So cool. Reminds me, I should learn me a Haskell... for great good! https://learnyouahaskell.github.io/introduction.html
380.
by andai 2mo ago
So here's the workflow I gathered, from the comments I've read here recently: - Claude as main agent, but use this skill[0] to make Claude delegate everything to Codex, because it's cheaper and faster. (Hilariously, the skill
381.
by andai 2mo ago
I noticed with with OpenAI's reasoning models (o3, o4-mini), and early GPT-5 (but they fixed it there, at least in chat). It went from the 4o "over-familiar" sycophancy to sounding like an absolute robot. I think it's be
382.
by andai 2mo ago
>Part of the pleasure of art, is making art. Yes! I'll have to write a response to this article... an ode to crap art! A while back I started consciously making low effort art. It was part of a exercise where I was intentionally des
383.
by andai 2mo ago
It's a sales problem. "Why should I care about this in the first place?" It doesn't necessarily have to be for something, but if you can't communicate any value, then... yeah... the results speak for themselves.
384.
by andai 2mo ago
When I read something and don't understand it, is it because the author is financially incentivized to make it sound more impressive than it is, or because they're unable or unwilling to communicate clearly, or both? Sometimes! Of
385.
by andai 2mo ago
Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away. —Antoine de Saint-Exupéry
386.
by andai 2mo ago
Well, a teacher's job is to help illuminate a cave. I can grope around blindly just fine on my own!
387.
by andai 2mo ago
Father, I cannot zoom the paintings! Edit: found a zoomable version! https://archief.ntr.nl/tuinderlusten/en.html
388.
by andai 2mo ago
You can find people who will do it for free, there are forums dedicated to that. (Called accountability partner, accountability buddy, things like that.) I set up a project with a friend once, where every evening, for several months straigh
389.
by andai 2mo ago
I am very clumsy, and I've become excellent at catching things mid fall!
390.
by andai 2mo ago
The real fun one is what do we do when civilization breaks. Do we even remember how we assembled it in the first place? I don't mean things like plumbing and roads, I mean things like, "what happens to a culture when people stop s

More ›