youoy

8 ms·

61.
by youoy 2y ago
An ugly proof is super useful. It turns a statement into a theorem. There is a famous quote by Riemann: "If only I had the theorems! Then I should find the proofs easily enough. " Once you have a proof, simplifying it should be mu
62.
by youoy 2y ago
It's a fundamental problem with using AI to do "intuitive" math, but not a fundamental problem with AI to do formal math. As you have stared, a lot of mathematics can be formalized. For me the problem for AI with math is goin
63.
by youoy 2y ago
Hmm... I thought every ML algorithm was already intelligently self-modifying software. I would love if the release was just a linear regression optimization algorithm :)
64.
by youoy 2y ago
I love this one: > My mother came in with a great tip: when in doubt, ask second time parents, not other first timers. So true in so many levels.
65.
by youoy 2y ago
I unexpectedly went through that when my second kid was born. I somehow felt that my first kid grew up suddenly and I was not able to say goodbye to the little boy that he was while a single kid.
66.
by youoy 2y ago
Of course it's not like this for everyone. Even for him it is not like this all the time :) but if you listen to the world around you there are these bits of beautiful life that appear here and there. He says this at some point: > T
67.
by youoy 2y ago
> The training pipeline will be: raw video -> Genie 2 -> game engine with rules -> physical robot One of those arrows is not like the others
68.
by youoy 2y ago
"Everyone lies one day. Everyone. Even dads. But not on the internet. Not HN. HN won't lie" – My granpa, 86
69.
by youoy 2y ago
It depends if you want to be able to prove new things by yourself or not. If you want to do it, then you definitely need to understand /recall all of the whys of every section of the proof. They are all there for a reason. If you don&#
70.
by youoy 2y ago
In my case I realized I achieved a whole new level when I tried to do all of the proofs by myself. I used to read the whole subject a few times and after that, I would go one definition/proof at a time and first to recall the statement
71.
by youoy 2y ago
> Ads only get viewed as long as Search is high quality and people don't switch to a competitor. This is what someone thinking long term would conclude. On the contrary, lowering the quality of your search causes people to spend mor
72.
by youoy 2y ago
I agree with you! Correlation is not causation. Being particularly motivated about a subject causes you to read more books, and both combined cause you to improve on that subject. Setting the goal to arbitrarily read more books just because
73.
by youoy 2y ago
Both approaches are not incompatible. It's probably more efficient to build a high level map of the subject using podcasts/YouTube videos than reading a dense book. Once you have that high level map, you have the tools to choose t
74.
by youoy 2y ago
Ah, I was not aware of that! Thanks!
75.
by youoy 2y ago
I do the same! It works pretty good for "visually addictive" apps... but not for HN for now
76.
by youoy 2y ago
> If a US school year is 9 months and the summer vacation is 3 months, it means that every 3 years, a student who only does math in the summer will be accelerated by one year. This means that after 9 years of school with math in the summ
77.
by youoy 2y ago
Is it really marketing in general? I can agree with some of it, but for me the existence of the term "low hanging fruit" to describe some statements says otherwise...
78.
by youoy 2y ago
I would say that "narrow" mathematics (finding a proof of a given statement that we suspect has a proof using a formal language) is much easier that "generally" laying brick or cutting hair. But I cannot see how consiste
79.
by youoy 2y ago
What do you mean by that? We have the monosemanticity results [0] [0] https://transformer-circuits.pub/2024/scaling-monosemanticit...
80.
by youoy 2y ago
It's more like looking at grided paper and thinking that defining some rules of when a square turns black or white would result in complex structures that move and reproduce on their own. https://en.m.wikipedia.org/wiki
81.
by youoy 2y ago
Don't get caught in the superficial analysis. They "understand" things. It is a fact that LLMs experience a phase transition during training, from positional information to semantic understanding. It may well be the case that
82.
by youoy 2y ago
I actually think that this is the most important part of the article: > Similarly, it’s important to be open about how you use ChatGPT. The simplest way to do this is to generate shareable links and include them in your bibliography . B
83.
by youoy 2y ago
I think this article marks the "peak of inflated expectations" of AI for HN posts.
84.
by youoy 2y ago
From the conclusions: > I find that AI substantially boosts materials discovery, leading to an increase in patent filing and a rise in downstream product innovation. However, the technology is effective only when paired with sufficiently
85.
by youoy 2y ago
I was not aware of the additive noise part. I will have to look into that, thanks for the info!
86.
by youoy 2y ago
From the linked article: > The method, in the form of an algorithm, takes in data that have been collected over time, such as the changing populations of different species in a marine environment. From those data, the method measures the
87.
by youoy 2y ago
Fair point. I mean, if you compare any philosophical statement against a mathematical one, the philosophical one is always going to be almost empty in comparison :)
88.
by youoy 2y ago
I share the translated quote that they have chosen for the frontpage, for people that just read the comments: > There are no “final words”, no “conclusions” in Récoltes et Semailles, any more than there are in my life, or in yours. There
89.
by youoy 2y ago
Not to mention that math proofs are more than graph trasversals... (Although maybe simple math problems are not) There is the problem of extracting the semantics of math formalisms. This is easier in day to day language, I don't know t
90.
by youoy 2y ago
Why do people still insist that this is unlikely? Like assuming that the company that payed 15M for chat.com does not have some spare change to pay some graduate students/postdocs to solve some math problems. The publicity of solving s

More ›