Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
abetusk
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
301.
▲
by
abetusk
2y ago
I forgot to mention that Kenney has a cult following. Reddit often labels him as "Asset Jesus" [0]. It looks like there's even a site dedicated to Asset Jesus [1]. [0] https://www.reddit.com/r/ThemeParkit
302.
▲
by
abetusk
2y ago
Kenney is awesome. OpenGameArt.org (OGA) has a lot of libre/free assets (Kenney often posts on OGA): https://opengameart.org/ Itch.io also has many CC0 and CC-BY licensed assets: https://itch.io/game-as
303.
▲
Dimensional Reduction and Topological Recovery from Neuronal Spike Data
(twitter.com)
1 points
by
abetusk
2y ago
|
1 comments
304.
▲
by
abetusk
2y ago
Original paper: https://winnieyangwannan.github.io/RippleTagging/Paper/
305.
▲
by
abetusk
2y ago
And there's a generalized Hilbert curve, the Gilbert curve, for non powers of two rectangular regions [0] (online demo [1]). [0] https://github.com/jakubcerveny/gilbert [1] https://jakubcerveny.github.i
306.
▲
LumonOS
(strawstack.github.io)
3 points
by
abetusk
2y ago
|
1 comments
307.
▲
by
abetusk
2y ago
Video: https://www.youtube.com/watch?v=C3lXoiiYxRY GH: https://github.com/strawstack/TheFourTempers
308.
▲
by
abetusk
2y ago
I agree. This is also a lack of effort on the journal's part to set expectations of what the reviewers should be looking for in an accepted paper. In the journal's defense though, what most likely happened is that the reviewers we
309.
▲
by
abetusk
2y ago
The second post in that thread is gold: """ ... I once almost solved a conjecture, establishing the result with an "epsilon loss" in a key parameter. We submitted to a highly reputable journal, but it was rejected
310.
▲
by
abetusk
2y ago
For the lazy, the blog post that accompanies it (2012): http://bit-player.org/2012/world3-the-public-beta The source code: https://github.com/bit-player/limits I haven't read "The Limits
311.
▲
by
abetusk
2y ago
> ... i think google will take the lead here as they just have monopoly on dataset ... I think there's plenty of data for people to train substantial models, either drawing directly from the public domain or scraping content. The co
312.
▲
by
abetusk
2y ago
Musk's net worth was $137B in 2022, $270B in 2023 and now estimated to be $400B. $1T is a bit of an extravagant prediction but it's within the realm of possibility. And it might not be him. Bitcoin was around $43k at the end of 20
313.
▲
by
abetusk
2y ago
* Solar energy will account for around 8% (give or take 1%?) of the worlds energy usage * We'll see the worlds first trillionaire * Bitcoin will reach $200k+, and remain largely stable around that price, at the end of 2025 * Generative
314.
▲
by
abetusk
2y ago
I have a few of the Manga Guide books and I haven't found them so great for the way I learn. The Manga Guide books are in a kind of "Plato's dialogue" style which I find off-putting but, for people who like ideas present
315.
▲
by
abetusk
2y ago
"The Manga Guide" [0]. Japanese Python book and other related manga programming books [1]. [0] https://nostarch.com/manga-guide-13-book-set [1] https://www.amazon.co.jp/12%E6%AD%B3%E3%81%8B%E3%82%8
316.
▲
by
abetusk
2y ago
Let's say I'm a huckster that plays a game with you where if I roll a single six sided die and it lands on 1 you lose but you win otherwise. Let's say you have some guarantee that I'm using the same die each time and tha
317.
▲
January 1, 2025 is Public Domain Day: Works from 1929 are open to all
(web.law.duke.edu)
3 points
by
abetusk
2y ago
|
0 comments
318.
▲
by
abetusk
2y ago
Here's a more pithy summary: "We've made a copy of the internet, run current state of the art methods on it and GPT-O1 is the best we can do. We need better (inference/search) algorithms to make progress"
319.
▲
by
abetusk
2y ago
I'll give my take: * Before the current renaissance of neural networks (pre ~2014ish), it was unclear that scaling would work. That is, simple algorithms on lots of data. The last decade has pretty much addressed that critique and it&#
320.
▲
by
abetusk
2y ago
Do you mind going into what web stack you use? What about businesses that want online payments?
321.
▲
by
abetusk
2y ago
From the EULA.txt [0]: """ ... DESCRIPTION OF OTHER RIGHTS AND LIMITATIONS You may not reverse engineer, decompile, or disassemble the SOFTWARE PRODUCT, except and only to the extent that such activity is expressly permitted
322.
▲
Gaussian Opacity Fields Efficient Adaptive Surface Reconstruction
(niujinshuchong.github.io)
2 points
by
abetusk
2y ago
|
0 comments
323.
▲
by
abetusk
2y ago
Again, not an expert but Dr. Sarah Taber claims that number is inflated by considering farm employees, not farm owners [0]. The claim is that farm owners are no worse than the baseline but that farm employees, often times low wage, migrant
324.
▲
by
abetusk
2y ago
I certainly am no expert but Dr. Sarah Taber talks about how many farms, especially small to mid sized, are effectively tax write offs and that the "put upon" farmer is a myth [0] [1]. [0] https://open.spotify.com/
325.
▲
by
abetusk
2y ago
Could you go into more detail?
326.
▲
by
abetusk
2y ago
Strava is a fitness app. So, apprently, the security detachment of political figures tends to use the app, presumably because they're into fitness and keep in shape, and their location can be tracked through the app. As the security de
327.
▲
by
abetusk
2y ago
Good article but some points I think it misses: > "Case, the hapless protagonist, stumbles between crises, barely knowing what’s going on, at risk from a femme fatale and being made offers he cannot refuse from mysterious Mr. Bigs.”
328.
▲
by
abetusk
2y ago
The article goes into (what I consider) actionable methods. Specifically: * Create a finance department that's independent in both their reporting and ability to confirm metrics reported by other departments * Provide a periodic meetin
329.
▲
by
abetusk
2y ago
There was someone else who linked to a longer talk from Sannikov where he directly addresses specular GI: https://www.youtube.com/watch?v=TrHHTQqmAaM&t=2275s
330.
▲
by
abetusk
2y ago
Am I correct in thinking that this approach effectively makes real time GI a solved problem? Can someone who's more knowledgeable than myself offer any explanation on why it took so long to discover a method like this? I remember heari
More ›