killerstorm

5 ms·

31.
by killerstorm 22d ago
Xanadu principle have outlined important problems. This does not mean that Xanadu created practical solution, though: it seems they were focused on text transclusions UI for some reason... If the goal of Xanadu was to offer a better, more p
32.
by killerstorm 23d ago
Something like that. I think DNS registry system works pretty well under ICANN supervision: user isn't tied to a specific registrar.
33.
by killerstorm 23d ago
The government haven't noticed that email became as important as regular mail, unfortunately.
34.
by killerstorm 23d ago
That's just over-fit. Happens in humans too - like an alcoholic might swear to stop drinking. It was demonstrated that LLMs are capable of non-trivial self-introspection. E.g. if a steering vector is injected into residual stream, a su
35.
by killerstorm 23d ago
Perhaps "true" self-referentiality was not needed. But it seems "dynamic feedback" is essential, and it seems to be adjacent to self-referentiality. If we look at optimization process, it first does a forward pass which
36.
by killerstorm 23d ago
A lot of language nuance can only be picked up through cultural immersion, e.g. seeing how native speakers interact. (Which is a lot easier nowadays via Netflix, Youtube, etc.) It's pretty much impossible to "speak your language w
37.
by killerstorm 24d ago
Eh? STE is designed for procedural writing, e.g. "if you have a problem, press this button". I generally don't ask advanced AI to give me step-by-step instructions.
38.
by killerstorm 25d ago
This video actually explains math reasoning on a high level. Have you actually watched it?
39.
by killerstorm 26d ago
Rationalists generally prescribe a 'Bayesian' world-view, which can extract useful information out of a chaotic, too-complex-to-model world. Regarding the power of intelligence, it's generally considered to be synonymous with
40.
by killerstorm 26d ago
What "sci-fi baggage"? "Intelligence explosion" was first described by I.J. Good, a mathematician. von Neumann described singularity as a result of accelerating technical progress. He's also a mathematician, not a s
41.
by killerstorm 26d ago
When the "LessWrong crowd" talks about dangers of AI, they don't assume a particular form of intelligence or method to achieve it. They talk about the danger of optimization processes, i.e. "find X which minimize Y(X)&qu
42.
by killerstorm 26d ago
Oh, where are the AI unsafety organizations to provide us with truly unbiased investigations...
43.
by killerstorm 1mo ago
Fun fact: in Soviet Union a common drug which people believe might reduce chance of a heart attack or stroke in high-stress/panic situation was Validol. What people don't know is that it was just peppermint oil + glucose. I.e. a m
44.
by killerstorm 1mo ago
This remind me kind of stuff people did back in 2023: https://x.com/SigGravitas/status/1642181498278408193 ... Massive Update for Auto-GPT: Code Execution! Auto-GPT is now able to write it's own code using #g
45.
by killerstorm 1mo ago
"The bigger they are, the harder they fall." If you're smart you probably have bigger, more challenging goals, so chances of success aren't higher. Deliberately setting goals lower might not help. Having a mindset which
46.
by killerstorm 1mo ago
MIDI is still widely used for professional music production. It sounded goofy back in the day because synthesizer it was played on was not very good.
47.
by killerstorm 1mo ago
Suppose you came up with AI algorithm which is better than transformers and such. Your options are: 1. publish it e.g. as a paper or open source, letting all big AI labs to use it for free 2. keep it secret Are you saying that it's evi
48.
by killerstorm 1mo ago
> the guy in first place has to spend all that time and effort to develop the thing The problem is that patenting something doesn't require one to put a lot of time and effort - just to produce few pages of text and diagrams. So it
49.
by killerstorm 1mo ago
The question is whether it can accelerate scientific research, not whether it can completely replace human creativity.
50.
by killerstorm 1mo ago
Doubt. My daughter wanted to make a static web to publish illustrated stories she made an such. I happened to have a Loveable subscription I forgot to cancel so I figured she can at least prototype it there. Well, it did the web site alrigh
51.
by killerstorm 2mo ago
I'm not saying "web" would be better if it was more centralized. WWW is optimal in the sense that it sits directly on top internet protocols, i.e. it just defines how to use TCP/IP to retrieve a document. I think in the
52.
by killerstorm 2mo ago
Hmm, that's basically a fancy way to describe psychological pressure and persuasion.
53.
by killerstorm 2mo ago
I don't think it's really an "AI problem", we just got to the "worse" part of the "Worse is better". Back in the day one of competitors of the World Wide Web was Project Xanadu. Project Xanadu was sup
54.
by killerstorm 2mo ago
Are we sandboxing AI agent harness process, or the environment it executes commands in? Ideally, they should run in _different_ sandboxes. The environment might corrode the harness (e.g. rogue npm/pip packet would manipulate agent harn
55.
by killerstorm 2mo ago
What if it puts malicious code into test file and you allow `npm run test `?
56.
by killerstorm 2mo ago
There are more than two types. Or, at least, more than one way to tinker with the code. Just like aleph_minus_one, I like to tinker at a higher level: find a more general way to solve things, find a new way to structure abstractions, etc. &
57.
by killerstorm 2mo ago
In a short story titled 'Glory' Greg Egan described process where hydrogen and anti-hydrogen are forged together into a thin needle: > The end point of this process of nuclear sculpting was a sliver of compressed matter and ant
58.
by killerstorm 2mo ago
Google's revenue is 10x up in the last 10 years.
59.
by killerstorm 2mo ago
What's the obvious limitation? To Yann LeCun it was "obvious" that you can't build a world model from a text, so he predicted that even "GPT 5000" can't predict that an object sitting on the table would mo
60.
by killerstorm 2mo ago
This is literally an opinion of one dude which is not backed by any kind of quantitative evidence. It's actually possible to answer this question rigorously: 1. Define a scientific result which qualifies as a "jump". They sho

More ›