by selling me a license that lets me run their software on my own machine, not theirs. Like in the old times

going to “C:\Users\user\Documents” in explorer, vs just typing in “documents”. One takes you to your documents folder, which will be empty, the other takes you to some other path from onedrive

on some sites the plugin fails to properly detect which fields correspond to which, true (usually when javascript fuckery is involved). But fixing that by manually pointing out the fields once on such sites is easy enough for me. I also switched firefox to use keepassxc for passkeys, which makes them actually portable and usable for me.

asking such an open ended question doesn’t mean much when nowadays, more and more people consider “anything I don’t agree with” to be hate speech.

what sucks about keepassxc?

a > 30000 vehicle that won’t be available for a decade

matrix multiplications. lots and lots of matrix multiplications. What gpus are good at already

the exact same intended use case, in fact

not really. A lot of techniques have been known for decades. What we didn’t have back then was insane compute power.

and there’s the turing award for computer science.

and physicists use tools from math, so fields medals should be awarded to physicists.

playing nintendo games, on a pc, using a dualshock feels so wrong and yet so right

anyone remember the anarchist’s cookbook?

randomly doesn’t mean equiprobable. If you’re sampling a probability distribution, it’s random. Temperature 0 is never used, otherwise a lot of stuff would consistently hallucinate the exact same thing

if it’s allowed to use its own interactions as data, it will collapse. This has been studied. Stuff just does not work the way you think it does. Try coding one yourself.

The llm does not give you the next token. It gives you a probability distribution of what the next token coould be. Then, after the llm, that probability distribution is randomly sampled.

You could add billions of attention heads, it will still have an element of randomness in the end. Copilot or any other llm (past, present or future) do have this problem too. They all “hallucinate” (have a random element in choosing the next token)

turning jhonny into an llm does not work. because that’s not how the kid learns. kids don’t learn math by mimicking the answers. They learn math by learning the concept of numbers. What you just thought the llm is simply the answer to 2+2. Also, with llms there is no “next time” it’s a completely static model.

yeah. what’s your point. I said hallucinations are not a solvable problem with LLMs. You mentioned that alpaca used synthetic data successfully. By their own admissions, all the problems are still there. Some are worse.

from their own site:

Alpaca also exhibits several common deficiencies of language models, including hallucination, toxicity, and stereotypes. Hallucination in particular seems to be a common failure mode for Alpaca, even compared to text-davinci-003.

here’s that same conversation with a human:

“why is X?” “because y!” “you’re wrong” “then why the hell did you ask me for if you already know the answer?”

What you’re describing will train the network to get the wrong answer and then apologize better. It won’t train it to get the right answer

Yeah that implies that the other network(s) can tell right from wrong. Which they can’t. Because if they did the problem wouldn’t need solving.