pcwelder

12 ms·

121.
by pcwelder 3y ago
> The oracle problem is a problem because there are no oracles To expand on this and correct your first statement, physical engagement can't be a proof of personhood either. There can never be a proof of personhood, or if someone is
122.
by pcwelder 3y ago
Relevant discussion https://www.lesswrong.com/posts/HBHsFCeAMSqMD7Td7/does-the-s... I wondered aren't the structures where there's more randomness than what we see more likely since there are a lot of di
123.
by pcwelder 3y ago
> Supposing there's no ultimate, objective, metaphysical thing called meaning I disagree that there can't be a rationally derived objective utility function (purpose of life) that we can assign to ourselves. For one thing, we c
124.
by pcwelder 3y ago
So you proved to yourself that you can do hard things, what then? You don't need to continue proving the same, doesn't then that stop being a fuel for motivation? That's what happened to me at least.
125.
by pcwelder 3y ago
Would Chatgpt help in accelerating the learning process? Does it understand Nix well enough to be useful?
126.
Lemmy.world (and some others) were hacked(lemmy.world)
5 points by pcwelder 3y ago|1 comments
127.
by pcwelder 3y ago
So the law of increasing entropy is not a fundamental law of the reality because it can be derived from other fundamental equations. Suppose I show you a snapshot of a random universe, would you be able to tell if the entropy of the univers
128.
by pcwelder 3y ago
This is eye opening. Thanks a lot for this comment and linking the pdf. I loved E.T. Jayne's Probability Theory book, so looking forward to reading this pdf too.
129.
Why pathlib.Path doesn't inherit from str in Python (2016)(snarky.ca)
2 points by pcwelder 3y ago|0 comments
130.
by pcwelder 3y ago
>It looks too spaced out People have similar complaints with Reddit re-design (so do I). What's with making websites "prettier" instead of more usable?
131.
by pcwelder 4y ago
Neat visualisations and very erudite article. Thanks! My observation is that the final formula is equivalent to two tetrahedra joined on a shared face. There must be a good way to manipulate that directly to get the constant sums. I can fin
132.
by pcwelder 4y ago
This has been the single biggest learning for me as well, but in slightly different flavours: 1. In machine learning. For example, suppose you want to generate an article. If you try to build a model that sequentially generates all the word
133.
by pcwelder 4y ago
I don't want to circumvent the ban by creating a new account, it's against the rules, and mods will ban it without considering anything else. I could try convincing the mods, but it'll not be easy. > Sounds like you got th
134.
by pcwelder 4y ago
I have an opposite story. I created a reddit bot that recommended some words to the users on https://reddit.com/r/whatstheword subreddit posts. I could've done a lot of things to improve quality, but it looked goo
135.
by pcwelder 5y ago
"In the beginning the Universe was created. This had made many people very angry and has been widely regarded as a bad move."
136.
by pcwelder 5y ago
Pyantic and Cattrs both suffer with unacceptable issues and footguns. I have used Pydantic in production a lot and have played around with Cattrs. I've also looked at similar libraries like Dacite and marshmallow-dataclasses but none s
137.
by pcwelder 5y ago
https://github.com/samuelcolvin/pydantic/pull/1568 If I'm reading this issue right the benchmarks in the pydantic docs are misleading Also > The author of this article ( https://stefan.sofa-
138.
by pcwelder 6y ago
Your conclusion doesn't follow your premise. Whether it's still an apple is a question of categorisation and a human construct (1). But once it's established it is indeed an Apple you can count the number of such Apples. Coun
139.
by pcwelder 6y ago
Using a previous Stockfish scorer they trained an NN without any labelling effort. This is also similar to how unsupervised translation is done in some methods. They start from word->word dictionary results and iteratively train lang1-&g
140.
Meaning = Information and Evolution (2016)(arxiv.org)
1 points by pcwelder 6y ago|0 comments
141.
by pcwelder 6y ago
Also related to the fallacy of false dichotomy. https://en.wikipedia.org/wiki/False_dilemma
142.
by pcwelder 6y ago
Sometimes you don't need to have generative knowledge. I can recognise dog and distinguish it from a cat. Tell me to draw one and you'll get a degenerate outline which only remotely resembles a dog. For all my purposes I don'
143.
by pcwelder 7y ago
Thank you for pointing this out. Every time they show road accident statistics as an argument for the safety of self-driving cars or airlines, I wonder how is the risk conditional on me in the vehicle. Surely the risk substantially decrease
144.
by pcwelder 7y ago
I don't see his point. Doesn't renormalizing token counts essentially eliminate the effect of tokenization? The perplexity which then we get essentially is representative of how well a model compresses the test document. Isn'