voidhorse

7 ms·

61.
by voidhorse 2mo 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
62.
by voidhorse 2mo 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
63.
by voidhorse 2mo 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
64.
by voidhorse 2mo 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
65.
by voidhorse 2mo 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
66.
by voidhorse 2mo 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.
67.
by voidhorse 2mo 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
68.
by voidhorse 2mo 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
69.
by voidhorse 2mo 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
70.
by voidhorse 2mo 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.
71.
by voidhorse 2mo 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.
72.
by voidhorse 2mo 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
73.
by voidhorse 2mo 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
74.
by voidhorse 2mo 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
75.
by voidhorse 2mo 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
76.
by voidhorse 2mo 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,
77.
by voidhorse 2mo ago
#3 just means the person is going to stop asking you when they realize that: a. You bring no expertise beyond asking an LLm for the answer. b. They can just do that themselves. If someone is asking you something and you tell them "th
78.
by voidhorse 2mo ago
I think there is a clear distinction to make in the case of communications. Yes, if you are writing a longer piece of communication and you just verbatim ship whatever the LLM spits out, it is important that you note that in the communicati
79.
by voidhorse 2mo ago
Pragmatically, the range of application for a fuzzer and for an LLM are so widely different that I'm not sure it makes any kind of pragmatic sense to compare the two. Also, I don't hear a lot of people credit fuzzers. I do hear a
80.
by voidhorse 2mo ago
I completely agree. All crediting LLMs does is give the companies that built them precisely the kind of ego stroking feedback they want. I bet Dario giggles with excitement every time he sees "Claude" as the author of a git commit
81.
by voidhorse 2mo ago
I don't have a problem with Diataxis itself but every time it comes up it saddens me because it is a reminder of how bad the technical writing community (by which, here I mean people who have technical writer as a job title) has been a
82.
by voidhorse 3mo ago
Yeah, this letter is just proof of (the by now long standing fact) of corporate power/corporate capture of American governance. Tons of popular protests against big AI? Makes no difference. Joint corporate letter about big AI? Possibly
83.
by voidhorse 3mo ago
Yes because the keystrokes aren't free and the world is finite. We need to pay these monkeys in bananas.
84.
by voidhorse 3mo ago
Probably because claude suggested some kind of wack react based setup early on (because react dominates the training data) and it's be blasphemy worthy of termination for the Anthropic employees to question the sacred pronouncements of
85.
by voidhorse 3mo ago
> to be more specific, i'd argue LLMs “understand” awfully similarly to a brilliant (polymath) with early dementia or Alzheimer' At which point i'd argue that you're kind of refuting your own claim as most humans are
86.
by voidhorse 3mo ago
I'm really starting to tire of people making broad, general claims about how LLMs work or how to use them with N = 1 or 2. An LLM is a statistics machine for goodness sake. Basically any general claim about them needs to exploit the
87.
by voidhorse 3mo ago
I don't know enough about the quantum level to really understand whether or not that would change my view of things. From Shannon's theory, it's abundantly clear that "information" is just a mathematical description
88.
by voidhorse 3mo ago
What a stupid list. Heavily biased. I'd just like to point out that a large bulk of these papers depend on the existence of the computer and yet Alan Turing is not even on the list .
89.
by voidhorse 3mo ago
You are missing the point. Let me use "fabulashinaism" which is a theory I just made up that posits that the universe is made up of little gnomes called atoms. There is a sense in which all of my observations hence will be "t
90.
by voidhorse 3mo ago
Let me try again. We use miles to measure distance. Hopefully you agree with me that, miles, being a system of measurement to express relationships of distance, have no ontological existence themselves. I would find it completely bizarre,

More ›