voidhorse

36 ms·

31.
by voidhorse 26d ago
Why would amazon buy "crap"? Surely they want their model to succeed and they want to train it on valuable input, no? They have more than enough resources to determine whether or not the books are worth buying. They have been in b
32.
by voidhorse 26d ago
You don't know that. A major part of the problem here is the lack of transparency. Keep in mind we wouldn't even know this was happening were it not for investigative journalism.
33.
by voidhorse 26d ago
> This way, it will live forever instead, scanned and trained on, conjoined with the rest of our knowledge in a magic machine. But that's just it, it won't live on forever because, from the perspective of preservation, traini
34.
by voidhorse 26d ago
Then give up training on antiquated books. Why does an LLM aimed at providing utility for people living in 2026 need to be trained on rare (thus probably obscure) texts of yore in the first place? Because these companies have no real strate
35.
by voidhorse 26d ago
Yeah, which is completely fine. There's a major difference between: A. Company destroys a book forever for training. Its scan is locked away forever in company records. In this case: - This information is locked away in the improvisati
36.
by voidhorse 26d ago
That's not really the point. You are defending the right of a massive corporation, to harvest and systematically and unilaterally make it impossible for people to purchase such books in the future without any oversight, just so they ca
37.
by voidhorse 27d ago
Who gets to decide what's "special"? Written works might be personal stories or expression of feeling. I've read a lot of unknown old literature that has little to recommend it as far as literary history goes, but it all
38.
by voidhorse 27d ago
> So it reduces to a corporation selling services based on public-domain knowledge (I am ignoring the part where a miraculous advance is confined to a single unknown book). Lots of corporations do this, and there's nothing wrong wit
39.
by voidhorse 27d ago
What are we gaining other than your silly little hypothetical? The whole problem here is: - Not everyone agrees on the utility / value of AI. - A small handful of companies control frontier models and get to shape how these inputs are
40.
by voidhorse 27d ago
It's possible for someone to do something socially reprehensible and entirely distasteful and to be perfectly within the letter of the law while doing so. The article doesn't emphasize copyright issues or fair use, because that is
41.
by voidhorse 27d ago
Most people agree it isn't normal. The problem is the people at the levers of power think otherwise. Capitalism as it exists in the US today, with effectively no regulation and deeply intertwined with governance, but in the wrong direc
42.
by voidhorse 29d ago
I believed this idea of kids = dead dreams for a long time. I don't think it's true. I think the author of the blog is letting his particular interests and cherry-picked examples cloud his judgment. If you look beyond the narrow
43.
by voidhorse 29d ago
No I'm not. Russell spied the paradox in the Begriffscrift. One only needs to check the wikipedia page, even: > he announced the discovery to Gottlob Frege of the paradox in Frege's 1879 Begriffsschrift https://en.wi
44.
by voidhorse 29d ago
Hey man, I completely agree. I love the begriffschrift and recognize the very important role it played in the history of logic. The reason I say it's in the rubbish heap is because basically only experts and people with strange hobbies
45.
by voidhorse 1mo ago
Claude doesn't "interpret" anything. That's not how LLMs work. They are still sentence completing machines. You give input in, the model network fires the chain that produces the most likely related output. That's i
46.
by voidhorse 1mo ago
> The agent harnesses are an interesting topic. I believe they have large teams shipping a ton of changes weekly. In that environment, is it realistic to expect rock solid software with such a feature set to be developed in a few months
47.
by voidhorse 1mo ago
Why? Unjustified linguistic analogy is a huge part of AI companies success. If we want to actually understand how LLMs work and differ in their operation from humans we need to develop appropriate language, not latch onto some totally unpro
48.
by voidhorse 1mo ago
But can't you see how coming out with a strong statement like "i've been doing this, it works" before your app even has a single user and before you've had to contend with any externally filed bug reports and keep t
49.
by voidhorse 1mo ago
I completely agree. I have Bell's book on the infinitesimal approach and it is infinitely (hah) more intuitive (hah again) than epsilon-delta limit foundations. It trades a heady second order logical statement for simple algebra. There
50.
by voidhorse 1mo ago
I have a copy and like it much. However, i was always partial to Frege's Begriffschrift. His notation was really creative. It's a shame Russel's deflation of that project has sentenced it to the rubbish heap of history.
51.
by voidhorse 1mo ago
Eh, not all software engineering is made equal. To work on engineering real time communications systems, banking systems, or transit systems and low level infrastructure, you definitely need some chops. To design and build out core network
52.
by voidhorse 1mo ago
Yes, you're on the money. And this is a perfect example of how people should use llms. Use them to deepen your own research and understanding rather than use them to glaze and implement the first naive idea that comes into your head
53.
by voidhorse 1mo ago
The "Socratic Method" (aka maieutic) skills annoy me, precisely because when you read them they are the kind of low-effort, low-expertise crap someone who over relies on AI would naively come up with when tasked with the problem o
54.
by voidhorse 1mo ago
This. This is exactly how I use them and I have had no issues so far. I read the book myself, then I point the LLM at it to ask questions about notions I might be struggling with.
55.
by voidhorse 1mo ago
Ugh, you're right! I read too quickly. Please disregard me OP. I'll leave the comment here just in case it clarifies things for poor readers such as myself lol.
56.
by voidhorse 1mo ago
I would say start with software architecture books for large or highly distributed systems. This will get you thinking beyond code level to system component level with subject matter that's familiar. From there I would round things out
57.
by voidhorse 1mo ago
So, while it's true that your runs with high speed and distance when both are considered are Pareto points, your max speed run and max distance run overall are also Pareto points. So calling these high distance+speed runs "Par
58.
by voidhorse 1mo ago
Yes, the difference between coding (writing software) and engineering (making informed trade offs about system design) is effectively bound up in analyses like those supported by concepts like pareto efficiency. One should keep in mind thou
59.
by voidhorse 1mo ago
One nuance that people sometimes miss is that pareto optimality in the continuous case and discrete case are distinct. Using continuous case algorithms on discrete feasible set optimization problems will make you miss the interior optimal p
60.
by voidhorse 1mo ago
I don't think it's possible to eliminate cognitive debt when using LLMs, but you can definitely minimize it. Part of this means recognizing that you will only get, say, a 2x productivity boost instead of a mythical 100x increase,

More ›