js8

25 ms·

421.
by js8 1y ago
Good news though, we will still get Harrison Ford! He will definitely be a replicant.
422.
by js8 1y ago
I wish people tried to formalize what a "good taste" is in programming. Here's a simple programming language - combinatory logic with S and K combinators, and you can create a new combinator by naming any expression of combin
423.
by js8 1y ago
I found it amusing that someone would say something to the effect of "killing people is the furthest I am willing to go, but not further", as if there is any further... Anyway, I agree it is a verifiable fact, but it also can be a
424.
by js8 1y ago
It would certainly better support his statement that the people are on their best behavior if being monitored. If his statement is true, then the real Larry Ellison (not publicly known one) is worse than a genocide supporter. He basically d
425.
by js8 1y ago
'Yes,' said O'Brien, 'we can turn it off. We have that privilege.'
426.
by js8 1y ago
It probably already happened in the Anthropic experiments, where AI in a simulated scenario chose to blackmail humans to avoid being turned off. We don't know if it got the idea from the scifi stories or if it truly feels an existentia
427.
by js8 1y ago
How good are LLMs at actually playing the adventure games? Can they win?
428.
by js8 1y ago
The thing I miss the most in social networks is the ability to verify the provenance of information through social graph (chain of trust). Ideally, if I see a post or comment, a process that I control should be able to establish whether the
429.
by js8 1y ago
Hi, thanks! I can't access it but I have read some papers from Jonathan Seldin on Curry's illative approach, which seems to be the key here, e.g. https://people.uleth.ca/~jonathan.seldin/CCL.pdf However, I th
430.
by js8 1y ago
I really like combinatory logic and lambda calculus, and I think it should be used to define common semantics for various languages and computer systems. Lately, I was wondering about if it's possible to represent any logic by choosing
431.
by js8 1y ago
There has been several other events like that (Django, Pandas and data science..). I don't think Python's popularity can be ascribed to any single event, it just happens to be a language that is reasonably close to pseudocode with
432.
by js8 1y ago
> And Python got popular cause of LLM AI thing. Not really, Python got popular in 2000s because it was the only sane of the three choices - Perl, Python, TCL. You must be young.
433.
by js8 1y ago
Not true, have seen many CS papers that run experiments lately. And arguably experimental mathematics is a science.
434.
by js8 1y ago
I found this video https://youtu.be/IQvREfKsVXM interesting, especially because it mentions couple of AI studies/papers that argue in favor of much smaller (and more efficient) models. (And I have never heard of them.)
435.
by js8 1y ago
I downvoted you. Reason - I don't see/understand what argument are you making. It seems you're just stating the claim.
436.
by js8 1y ago
If every user of yt-dlp did as I do, then it would have exactly the effect that it needs to have. If yt-dlp is used by a small minority of users, why would Google be antagonistic to it? And if it's used by sizeable portion of users, th
437.
by js8 1y ago
I have YT Premium and if Google bans yt-dlp, I will cancel my subscription. I pay them not to do that.
438.
by js8 1y ago
No, I think we agree, that's why I wonder why you use the word "anti-american". And I think you should consider not using it, as it's inappropriate* to use that word to denote just critics of american government or ameri
439.
by js8 1y ago
Why are so worried, then, whether the world is anti-american or not? Why use that term, if you personally don't want see the world through nationalist optics? I saw a commenter the other day here claiming that in EU, there are lot of &
440.
by js8 1y ago
Most of the composite indexes of this kind are similarly meaningless. It's certainly not just the domain of NGOs. But to be fair, if it makes people to stop and look why they have different preconceived ideas, I think it's a posit
441.
by js8 1y ago
What I meant was similar to the article. If I have bit vectors of length 1000 bits (that will be the embedding). Let's say that every concept I want to model corresponds to a set choice of 10 bits being all set to 1, and the remaining
442.
by js8 1y ago
Let me give an optmistic counterpoint. We should go back to Gramsci, who observed that the regime holds itself to power mostly by ideological hegemony. If everyone is being wiretapped, it's difficult to construct and maintain ideology
443.
by js8 1y ago
I thought from the title this would be about who the UI. Take, for instance, Emacs. User owns the UI, can completely configure and script it, in fact, they're encouraged to. On the other side of the spectrum is something like a website
444.
by js8 1y ago
I dabbled in magic, and what I found beatiful about it is, as an audience, you're amazed once (because magician never does same trick twice). But as a performer, you're actually amazed three times! First time as an audience, when
445.
by js8 1y ago
Well, obviously, you don't believe what you see. :-)
446.
by js8 1y ago
It's not correct to call them orthogonal because I don't think the definition is a dot product. But that aside, yes, orthogonal basis can only have as much elements as dimensions. The article also mentions that, and then introduce
447.
by js8 1y ago
For example, 1010 and 0101 are orthogonal, but 1010 and 0011 are not (share the 3rd bit). Though calling them orthogonal is not quite right.
448.
by js8 1y ago
You can also imagine a similar thing on binary vectors. There two vectors are "orthogonal" if they share no bits that are set to one. So you can encode huge number of concepts using only small number of bits in modestly sized vect
449.
by js8 1y ago
I wish I knew better how to use them for these coding problems, because I agree with GP they're underutilized. But I think if you have constraint problem, that has an efficient algorithm, but chokes a general constraint solver, that sh
450.
by js8 1y ago
While I agree with the gist of the article, I think the AI example is poor, because we know AI can make stuff up and it's a problem. So this failure of AI to be reasonably correct weakens the argument. In the old days, you would rely o

More ›