Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mkl
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
40 ms
·
511.
▲
by
mkl
2y ago
I'd never heard of it. The "crystal mark" itself is a money-making thing for what seems to be a for-profit company: https://www.plainenglish.co.uk/services/crystal-mark.html
512.
▲
by
mkl
2y ago
That problem's a known-known for most people interested in any of this, and for anyone who's made it as far as the first picture.
513.
▲
by
mkl
2y ago
The degree-10000 polynomial is definitely overfitting - every data point has its own little spike. The truncated curves look kind of nice, but the data points aren't shown on those plots, and the curves aren't very close to them.
514.
▲
by
mkl
2y ago
Big discussion three months ago: https://news.ycombinator.com/item?id=42826536
515.
▲
by
mkl
2y ago
Fascinating to see this on HN! Crochet can be easy and fun, and is often very procedural/mathematical - in work like this there are building blocks that are chained together like function calls and loops. I wouldn't recommend st
516.
▲
by
mkl
2y ago
You're looking for early testers, but you already claim to have nearly 40,000 customer reviews. Your screenshot is AI generated. You somehow have an order of magnitude more movies than Netflix. You show text conversations that refer
517.
▲
by
mkl
2y ago
Yes, I was disappointed. The original paper https://arxiv.org/abs/2410.22046 did look at chord progressions. They also trained a machine learning model to predict the next chord. Some of the chord progression data
518.
▲
by
mkl
2y ago
They are questioning jancsika's assertion at the top of the thread that an interval can somehow contain more than two pitches: > An "interval" is the difference between two (or more) pitches.
519.
▲
by
mkl
2y ago
This is false: https://ai.google.dev/gemini-api/terms
520.
▲
by
mkl
2y ago
I tried to look it up. I think they didn't ever get FTL, based on https://www.reddit.com/r/threebodyproblem/comments/1blvikg/c... and https://forums.spacebattles.com/threads/tr
521.
▲
by
mkl
2y ago
Why are you not using an ad blocker? Ads are optional - I didn't see a single one.
522.
▲
by
mkl
2y ago
Faster than light travel isn't needed (and IIRC doesn't occur in the series the dark forest name comes from). Spectacularly devastating weapons are definitely possible - redirect an asteroid into an inhabited planet and you'
523.
▲
by
mkl
2y ago
Technically, it is: https://en.wiktionary.org/wiki/sentient (the first two definitions). You mean the science fiction definition though (the third), and it's because it's not self-aware and not very intellig
524.
▲
by
mkl
2y ago
Yes. The section itself says PDF differs significantly from PostScript. The required changes detailed there to transform a PDF to PostScript are substantial: add PostScript implementations of the PDF operators; extract and translate the p
525.
▲
by
mkl
2y ago
A normal frame rate is probably enough if you do it with groups of columns rather than a single column of pixels. That's what https://trains.jo-m.ch/ does with a Raspberry Pi camera, posted in https://news.
526.
▲
by
mkl
2y ago
That seems like saying "Python is basically a subset of C; just run the simple transformations Cython implements". PDF can be transformed into something a PostScript interpreter can understand in the same way Python can be transf
527.
▲
by
mkl
2y ago
> it's not really common knowledge that PDF is basically just a subset of postscript. Because that's not actually true? Check out the table in the PDF specification, Appendix A, p985, listing all the PDF operators and their to
528.
▲
by
mkl
2y ago
This was posted by the author a couple of weeks ago, but didn't get any traction: https://news.ycombinator.com/item?id=43504429
529.
▲
Breaking the Llama Community License
(notes.victor.earth)
144 points
by
mkl
2y ago
|
81 comments
530.
▲
by
mkl
2y ago
Not when the traffic is coming from 10s of thousands of IP addresses, with very few requests from each one: https://drewdevault.com/2025/03/17/2025-03-17-Stop-externali...
531.
▲
by
mkl
2y ago
The actual paper: https://academic.oup.com/mbe/article/42/4/msaf065/8107989 A better article about it: https://www.su.se/english/news/study-provides-snapshots-of-m...
532.
▲
by
mkl
2y ago
How well it generates images of other people's copyrighted/trademarked characters might be part of the obstacle.
533.
▲
by
mkl
2y ago
BTW Xe, https://xeiaso.net/pronouns is 404 since sometime last year, but it is still linked to from some places like https://xeiaso.net/blog/xe-2021-08-07/ (I saw "his" above and went lo
534.
▲
by
mkl
2y ago
It would be better to link to the main page, https://evidence.dev/ , which is titled "Evidence - Business intelligence as code". The current link would be better titled what it is, using the project name and the pa
535.
▲
by
mkl
2y ago
The boxes can run airgapped: https://cloud.google.com/blog/products/ai-machine-learning/r...
536.
▲
by
mkl
2y ago
Interesting that the hardware is NVidia Blackwell, not Google TPUs. That means Google will likely have an energy efficiency and cost advantage, and keep their proprietary hardware out of other people's reach.
537.
▲
by
mkl
2y ago
> Now we have evidence that both chimps and bonobos have syntax, it is inevitable that this capacity for compositionality was inherited from our last common ancestor, says Leroux. “They just showed, unambiguously, that this core building
538.
▲
by
mkl
2y ago
Have you tried ZeroTier? Their free plan's been working well for me. I haven't tried NetBird.
539.
▲
by
mkl
2y ago
> it’s odd to attack observable since everything they put out is fully open-source Really? They've finally open-sourced the notebook editor? I can't find it on their GitHub. The long-proprietary notebook editor is a big part
540.
▲
by
mkl
2y ago
That seems much more likely to be cheetah-ancestors than dire wolves.
More ›