Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
usgroup
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
35 ms
·
481.
▲
by
usgroup
4y ago
Well, I think the hope is that formulae reveal relationships. Notable applications of SR have mostly been to physics and other science phenomena. So, you do some SR to understand the mathematical form of the relationship, and then carry on
482.
▲
by
usgroup
4y ago
I'm not sure that this wasn't known before. They claim "...while some works have hinted at the possible hardness of SR, no proof has yet been given that SR is, in fact, NP-hard.". Here is a counter example: https:/
483.
▲
by
usgroup
4y ago
Non linear forms most importantly.
484.
▲
by
usgroup
4y ago
I live in the world of data lakes and elaborate pipelines. Now and again I get to use a traditional star schema data warehouse and … it is an absolute pleasure to use in contrast to modern data access patterns.
485.
▲
by
usgroup
4y ago
That’s reads like it took a lot of effort to write.
486.
▲
by
usgroup
4y ago
You should make a logical fallacy trainer app -- it'd be a service to humanity.
487.
▲
by
usgroup
4y ago
Sextant navigation
488.
▲
by
usgroup
4y ago
Is there a book you'd recommend in this vein?
489.
▲
by
usgroup
4y ago
Except it looks like China did lift 800 million people out of poverty: https://www.worldbank.org/en/news/press-release/2022/04/01/l...
490.
▲
by
usgroup
4y ago
This one? https://journals.plos.org/plosmedicine/article?id=10.1371/jo... Lol yes, research finding Yes, true relationship No is missing a left bracket in the denominator.
491.
▲
by
usgroup
4y ago
Belt is better still by the same arc of reason.
492.
▲
by
usgroup
4y ago
Useful context: “I firmly believe that in order to understand something completely, you have to build it on your own from scratch. I used to do gradient calculation analytically, and thought that autograd was some kind of magic. So this was
493.
▲
by
usgroup
4y ago
Seems like a much more dev centric community.
494.
▲
by
usgroup
4y ago
I think the fundamental reason this is true is because devs over engineer on purpose to make a meal out of simple things. Pretending that your thing will need all the bells and whistles means you can claim you did all sorts of glorious thin
495.
▲
by
usgroup
4y ago
did anything succeed UML without the bloat and problems? I liked UML in principle but it just never was very useful to me beyond drawing boxes and arrows and the occasional sequence diagram.
496.
▲
by
usgroup
4y ago
... so long as the unseen domain has substantially overlapping mutual information with the training set ...
497.
▲
by
usgroup
4y ago
That medium sized dry objects in photographs and the matter of satellite imagery (for example) have little in common strikes me as self evident. The separation of "lines", "colours", etc at the most abstract level is a t
498.
▲
by
usgroup
4y ago
Yeah exactly — deep learning became the universal hammer some time ago, and I think people forget that it is far more sparsely used and a whole less dominant outside fairly specialised domains such as text/image use cases.
499.
▲
by
usgroup
4y ago
I think there were fewer programmers and programming was applied to fewer things. I think over time both aesthetic variability, new entrants and the proliferation of corporate backed super frameworks (which carry needless complexity from da
500.
▲
by
usgroup
4y ago
Hikaru tried to answer just that: https://www.youtube.com/watch?v=qjtbXxA8Fcc
501.
▲
by
usgroup
4y ago
Well she shows 6+ games where he has 100% correlation with the engine. What are the chances?
502.
▲
by
usgroup
4y ago
Yeah this is interesting … she shows that Hans had many tournaments where he shows record setting move accuracy as measured by correlation to Stockfish 15, and that 6 of these tournaments occurred in a row. She also shows that for those tou
503.
▲
by
usgroup
4y ago
you’re assuming that a person in power has an objective that somehow compliments them acting in an accommodating way: plenty simply do not.
504.
▲
by
usgroup
4y ago
Tbh I read ops advise almost like an endorsement for a self defence course. Bigger, stronger people shouldn’t punch us in the face and take our things — but they sometimes do, so you should learn how to hold your own.
505.
▲
by
usgroup
4y ago
Feels like calculating forcing chess moves. I.e forced mates are conditional on where pieces can go and where the pieces actually are.
506.
▲
by
usgroup
4y ago
lol -- brilliant. Notice he was using Github Copilot -- that's what all those green blocks are. He' just got the 31337 CLI version.
507.
▲
by
usgroup
4y ago
Scan the inside of your fridge — list all the dishes you could make.
508.
▲
by
usgroup
4y ago
Toy project: I’m learning a bit of Rust. I’m making a text compression and printing algorithm to encode as much data as possible on a single A4 page. It’s basically Hoffman coding + optimising transformations + encoding to a subset of Unico
509.
▲
by
usgroup
4y ago
You should do this for HN book recommendations
510.
▲
by
usgroup
4y ago
I mean doesn’t heaving salaries and a sellers market imply that engineering management jobs are not bullshit? Are there any jobs which command big salaries and which companies hire for directly that are bullshit?
More ›