> 1 / 0
Infinity
> 1 / -0
-Infinity
> 0 === -0
true
> Object.is(0, -0)
false1,261 karma · joined May 6, 2011
> 1 / 0
Infinity
> 1 / -0
-Infinity
> 0 === -0
true
> Object.is(0, -0)
falseFor example, if you were to ask a human artist to draw this image with two crabs, and this artist didn't know how many legs crabs have, what would they do? I think they'd probably google how many legs crabs have, and how they look in general, and at the very least draw two crabs with the same amount of legs.
But these generative models don't know how many legs crabs have, and they are not clever enough to ask, or even count it seems. So you end up with results like these. The main crab has 8 legs (6 + 2 claws), while the little one seemingly has 10 (or at least it has 5 on one side), and it's also missing an eye for some reason.
A human artist can also make mistakes, for sure. But i don't think these kind of mistakes would be expected from an artist.
And i agree with the other commenters that said this reflects poorly on the author of the book. Didn't they care to check the cover for obvious mistakes? Attention to detail is a very important trait on a technical writer.
I don't really know. The first thing that came to my mind would be to ask to draw/model different cross-sections of a 4D object ("cross-volumes"?).
We can visualize 3D objects, and therefore can draw 2D cross-sections of 3D objects relatively well, and relatively easily. Like, sections of a human body, or a house. So, maybe someone who can visualize 4D objects in their head could also model 3D "cross-sections" of that object at arbitrary "cuts". And we could check if those 3D radiographies are accurate, because we can model those 4D objects on a computer, and draw their 3D cuts.
Just a simple idea. I'm sure there could be other ways of probing this.
These things don't seem to depend on each other. And different people want to do different things in this world. I don't think that the people who want to help shelter homeless people would be deterred by other people wanting to send a space probe, or vice-versa.
I picked housing from the problems you mentioned just because it was the first one. I think this argument would still apply for the others. The reasons we're having those problems have nothing to do with how many probes we send to space.
I drive on Uber part-time. In my city at least, rating is not such a big deal. The vast vast majority of people give 5 stars ratings (i'm talking about 99% of ratings).
I assume there are some people who only give ratings when the service was exceptionally good or exceptionally bad, or would otherwise consider 3 stars as a fair rating for a normal experience. Let's call there honest raters. As a driver, i don't think honest raters can have such a big impact on me, because the platform is designed so that there's no preference to always match honest raters to some particular drivers. Their honest ratings will be spread more or less uniformly among drivers.
And, if we want to apply Kantian logic and think "but what would happen if everybody did the same?", well, Uber is a live system too: if more people would become honest raters, then the ceiling of what's the minimum average rating you need to have as a driver would be lowered accordingly. Uber needs drivers too. It needs to keep a healthy balance of available drivers and available passengers. If either of the two gets too low, if would lose the other part in short time.
Saludos desde el otro lado del charco!
And, to contribute something to the discussion: yeah, here in Argentina we have a similar situation. Import taxes are stupid high for some (most) products, making them prohibitively expensive for most people. Although high levels of inequality allow some people to access these kinds of goods, so for exmaple there's a non-zero market for iPhones here.
In the case of EVs, i haven't seen any cars here yet[1] —i suspect because of total lack of infrastructure for them, besides the high prices— but there's quite a lively market of electric scooters in Buenos Aires. They are not that expensive, and people can charge them in their homes.
This Uruguayan initiative of retrofitting electric motors into old cars is encouraging. I hope regulations serve a good purpose like setting good standards, instead of killing the movement outright :/
[1]: There's an Argentinian-made electric car with which i unfortunately share a nickname: Tito. But it's a total scam; a Chinese design ripoff with a ridiculously high price for what it's worth.
It talks about how large portions of the software industry are engaging in a sort of simulation of actual work. The idea of "startup cosplay" is both funny and quite sad.
I found the article very interesting, i'd recommend you reading it :)
I've had dual-booted until a couple of years ago, when i decided to ditch Windows for good, since i wasn't playing videogames, and the few i may wanted to play were playable on Steam. If i had needed Windows for a specific videogame or program, i would still be dual-booting today.
For example, an API like `Element.prototype.addEventListener(string, callback)` wouldn't make much sense in Rust, IDK about Go. Sure, you can make it work, but you'll always be rubbing against an API that was not designed for that language. Hell, addEventListener isn't even on the Element prototype, but on the EventTarget prototype, which Element.prototype inherits from. And that's something else that wouldn't make sense at all in Rust: prototypical inheritance. But it's part of the DOM API. It's part of the web. So we'd have to deal with it too, even if we switch to a new language.
Does this make some sense? I feel like i'm repeating what's been said before in this comment chain, so maybe i'm missing something about what you're not understanding (:
So even if you get other languages working alright on browsers, they are always going to be second-class citizens IMO, because they are going to be interfacing with JS APIs.
I can't think of a way to avoid this TBH, other than designing a different API for the new language, at which point i think we'd be no longer talking about the web as we know it.
I guess for me, so far at least, some sites feel much more legit and human that the obviously bot-ridden mess that are the likes of Twitter/Instagram/FB. Like for example here or on Lobsters (and more on the latter) i have the feeling that it's mostly people talking with people. On the couple of relatively-small subreddits i visit, i feel the same too.
But i could be wrong of course. Maybe the tone of an HN poster is super easy for an LLM to copy; there's a reason why "shit HN says" exists after all. The only reason i have to believe otherwise is that, in comparison, Instagram or Twitter bots are so obvious and bland, and those companies have way more resources to throw at AI than HN or reddit :P
I mean us as in a network of trusted individuals.
For example, i've been appending "site:reddit.com" to some of my Google queries for a while now —especially when searching for things like reviews— because, otherwise, Google search results are unusable: ads disguised as fake "reviews" rank higher than actual reviews made by people, which is what i'm interested in.
I wouldn't be surprised if we evolve some similar adaptations to deal the flood of AI-generated shit. Like favoring closer-knit communities of people we trust, and penalizing AI sludge when it sips in.
It's still sad though. In the meantime, we might lose a lot of minds to this. Entire generations perhaps. Watching older people fall for AI-generated trash on Facebook is painful. I hope we acted sooner.
It's difficult to argue that arbitrary-precision integers are "simpler" (because they don't have an overflow edge-case on almost every operation) when they are so much more inconvenient to use.
It's funny how other programs don't seem to have this issue, and their users are able to learn new things without having to resort to an external manual. Makes one wonder about the design of everyday things!
thanks for the very thorough reply! it's fascinating to see the techniques used to improve LLM outputs :)
i'll reply to some specific points, but i think your main argument of trying to find ways of working effectively with LLMs is spot-on.
> try this yourself: generate a novel pangram without any sort of iterative process or revisions -- the first thing that pops into your head you must commit to paper. it's very hard.
yes, it would be indeed very hard. and i don't know why i would try to do it that way. notice also that i didn't instruct ChatGPT to do it that way either.
to generate a pangram, i'd probably start with some random phrase, count the letters i've used and which ones i'm missing, and then iteratively tweak the phrase to use more letters of the alphabet until i've used them all. that at least seems like a reasonable strategy. and i would expect any intelligent agent to do the same. not that particular strategy, but "the same" as in: to try to find a strategy that works. after all, isn't that a fundamental part of intelligence? being able to find solutions to novel problems.
i know that LLMs don't work that way. and that's fine. but that was also the main point i tried to make in the post: we're being sold LLMs as "intelligent", but they don't work in any way like what we would intuitively say it's intelligent.
So, why would we go so blindly for option 3 and risk it all by being the first ones to defect in an interstellar prisoner's dilemma? Especially if we have good reasons to suspect that they've known about us for long and have remained neutral so far. After all, Proxima Centauri is quite close, and our atmosphere gives ample signals of there being life here. And we haven't been particularly quiet either.
I haven't delved deep into this, so there might be some much more tight logic to it. But my first impression is that the dark forest hypothesis seems a bit forced; sort of constructed backwards in order to explain a cool idea (that space is full of civilizations but everyone is quiet).
For example, one of the assumptions it makes is that obliterating another space-faring civilization is easy. And, at the same time, that we (and everyone) have very little information of other civilizations. I don't see how these two assumptions can hold at the same time.
[1]: https://jekyll.github.io/minima/ [2]: the sidenotes are stolen from Tufte CSS: https://edwardtufte.github.io/tufte-css/
> It is, but it’s also a bit unsatisfactory, no? A lot of the time the particular statement is less important than the way it was obtained—why or how is something works seems much more relevant to me than the binary question of whether it is true. A novel result more often than not also uses novel proof techniques, which may then be adapted to other problems. One may learn a lot more from the proof than even from the applications of the result.
> (Perhaps the author does has a point about the different mindsets :))
And thanks for the video recommendation. Added to my watchlist :)
[1]: https://lobste.rs/s/vkqqvl/programming_vs_mathematical_curio...
I was thinking about generating a graph of connected nodes like the one you did. But TBH, i really like the ones on Wikipedia[1], and didn't really feel like making something of equal quality hehe.
[1]: https://en.wikipedia.org/wiki/Kaprekar%27s_routine#Kaprekar'...
I meant the post to be more of a personal story. Of what i, personally, wanted to explore when learning about the Kaprekar constant, and of realizing that my interest is more on the "programming" side of things (programming languages, "readable" code, etc) than the more pure maths side.
But i can see how the tone could make it seem that was a competitive division of sorts, especially the title. I'll try to take it into account for (hopefully) future posts.
For example, you can write something like:
bananas = 4
puts "you have #{bananas} banana#{'s' if bananas != 1}"
And this works because: `if` is an expression, `if` evaluates to `nil` if the condition is not met and there's no `else`, and `nil.to_s` is an empty string.> And logging / error handling / debugging seems like the context where you would least want nil to become an empty string
Yep, 100% agree. And i think Ruby's designers agree too, since the language has the shortest method possible, `p`, for print-debugging. You can do `p something` and that'll print a useful string for debugging (`""` for an empty string, `nil` for nil, `[]` for an empty array, etc).