scotty79

12 ms·

361.
by scotty79 3mo ago
I already have a small sub. I think I can still upgrade/extend it, which is great.
362.
by scotty79 3mo ago
OpenAI compaction is very different than Anthropic compaction, especially in 5.6 I use GLM5.2 and gpt-5.5 and 5.6 and never noticed any quality degradation near the limits of the context or due to compaction.
363.
by scotty79 3mo ago
I can't really blame them that the biggest labs focused on trainig and realeasing huge models. The niche for small models should be filled with medium sized labs doing distillations of the huge ones into consumer grade hardware runnabl
364.
by scotty79 3mo ago
> saying that Fable was barely an improvement on Opus. Like it or not, Anthropic have a genuine moat right now with that model, What's more interesting is that Anthropic moat shrunk to just that model. There's zero reason to us
365.
by scotty79 3mo ago
> were seen as the solution to all the problems of Internet communities And the solution turned out to be putting every user in their own safety bubble of recommended conten and expanding the bubble only when user gets adventurous. And
366.
by scotty79 3mo ago
I guess AI sycophancy does contribute to its success.
367.
by scotty79 3mo ago
ChatGPT was released in November 2022 so only the last decisive blow was dished by AI. Stack Overflow was already 40% below its peek when it was struck. However looking at the shape of this last drop, I think it wouldn't matter if the
368.
by scotty79 3mo ago
I think the reveal button is counted as an upvote for the book because it supposedly discovered my taste and it was based on th books I revealed, but not bookmarked.
369.
by scotty79 3mo ago
> judging happens via AI Why? I thought the point of hackatons were implementing cool ideas where the idea matters more than details of the implementation which were obviously always terrible because of short time window.
370.
by scotty79 3mo ago
I think AI is going to enable a new future where most software products will have userbase of 1. And then they are not going to suck.
371.
by scotty79 3mo ago
If anyone is interested in the text that flows around the cracks. Cracks are painted on canvas underneath and the text is laid out in spans with help of pretext javascript library. I asked gpt-5.5 on sub running via pilot-proxy in a custom
372.
by scotty79 3mo ago
similarly to how they are trivial for a human?
373.
by scotty79 3mo ago
How do humans get their results at these tasks? Don't they try until they succeed and claim their best run as their result?
374.
by scotty79 3mo ago
With this harness or without?
375.
by scotty79 3mo ago
> How many hours of your life would you say you’ve danced? I'd say less than 20? Maybe even less than that. Even counting short bursts, to especially compelling music, that did't last more than few seconds. The closest I ever f
376.
by scotty79 3mo ago
I'd bet gpt-5.6 could reimplement it in few days from this article.
377.
by scotty79 3mo ago
Each solution generates new problems. That's the general physics of intelligence in our universe. Also label of "fake" is completely arbitrary. Because someone might call cancer a fake problem, because people shouldn't l
378.
by scotty79 3mo ago
Why not something p2p and encrypted? https://peerloomllc.com/pearpetal/ No servers, no problem. You fear losing your device? I'm sure you have more than one, any additional one you have is a fully functional backu
379.
by scotty79 3mo ago
To me nearly all of the real world dancing looks like that. The dancers could literally be olympic level and I still don't see the connection between their movements and the sounds of the music. Music videos sometimes do sync for me bu
380.
by scotty79 3mo ago
I pretty much predicted this. If such smart models capable of doing math research fail so hard on such simple games the interface is the problem, not the model. Right harness provides a good interface between the problem and the intelligenc
381.
by scotty79 3mo ago
"The Coase conjecture predicts that a monopolist selling a durable good will be forced to lower prices quickly because consumers anticipate future price drops, undermining the monopolist’s ability to maintain high prices." That re
382.
by scotty79 3mo ago
Modern lifestyle seems to start at $700-$800 per month per adult pretty much everywhere and doesn't get much better even at few k.
383.
by scotty79 3mo ago
Will it have fast COUNT-s?
384.
by scotty79 3mo ago
> Now there are no more problems and no reason to run the AIs, Are you sure we are talking about the same species?
385.
by scotty79 3mo ago
I can't imagine a successful civilization where non-computing energy usage isn't a rounding error.
386.
by scotty79 3mo ago
Which is still great because it means neither of the two best financed labs in the world manage to produce even two models themselves that would beat Kimi K3.
387.
by scotty79 3mo ago
What I meant is, when statewide planned economy dies society gets impoverished, infrastructure and resources are destroyed on massive scale, some people die because of poverty and crime spike, many migrate to other countries with all ineffi
388.
by scotty79 3mo ago
I don't understand people who can't get past that. I just do the same what I do when humans compliment me. Gloss over sugary bs and focus on the content.
389.
by scotty79 3mo ago
Coding LLMs got good only very recently (about 3 months ago). And many people don't use the best ones. The amount of LLM generated bs dropped for me from about 20%, which made me very cautious, to nearly zero, where I see the agent do
390.
by scotty79 3mo ago
You are heavily self-selecting what part of humanity you interact in the real world as well. You pick your friends, places you live, places you frequent, times you choose to go out and so on. I didn't know how wide the humanity varianc

More ›