HackerLangs
TopNewTrendsCommentsPastAskShowJobs

stared

13,448 karmajoined il y a 14 ans
A curious being, doctor of sorcery. Posts, projects & resume at: https://p.migdal.pl/

Now: benchmarking AI at: https://quesma.com/blog/

Previously: co-founder & CTO at https://quantumflytrap.com/

Submissions

Urban heat island effect in Brussels during the late June 2026 heatwave

eu-space.europa.eu
5 points·by stared·avant-hier·0 comments

The true cost of saying "Hi" to an AI agent

quesma.com
9 points·by stared·avant-hier·4 comments

[untitled]

1 points·by stared·il y a 3 jours·0 comments

Qwen 3.6 27B is the sweet spot for local development

quesma.com
1,192 points·by stared·il y a 12 jours·759 comments

Show HN: Tree, truth, druid and tar share one Proto-Indo-European root

p.migdal.pl
5 points·by stared·il y a 16 jours·1 comments

Astro 7.0

astro.build
2 points·by stared·il y a 18 jours·0 comments

Show HN: Tree, truth, druid, dryad, tar and dendrite share the same PIE root

p.migdal.pl
2 points·by stared·il y a 18 jours·0 comments

Wolves are reconquering Europe. Can people learn to live with them?

science.org
46 points·by stared·il y a 18 jours·66 comments

Opus Magnum Bench

opusmagnumbench.com
3 points·by stared·il y a 22 jours·0 comments

Statistical detection of systematic election irregularities (2012)

pnas.org
2 points·by stared·il y a 24 jours·0 comments

GLM 5.2 ranks #2 in Code Arena: Frontend

twitter.com
2 points·by stared·il y a 24 jours·1 comments

Deepfakes are everywhere. The godfather of digital forensics is fighting back

science.org
4 points·by stared·il y a 26 jours·0 comments

Surpassing Frontier Performance with Fusion

openrouter.ai
3 points·by stared·il y a 26 jours·0 comments

Bye-bye Butterflies

europeancorrespondent.com
2 points·by stared·il y a 30 jours·0 comments

Can AI answer open questions in physics?

m-malinowski.github.io
4 points·by stared·le mois dernier·0 comments

Microbe with tiny genome may be evolving into a virus (2025)

science.org
3 points·by stared·le mois dernier·1 comments

Ireland paid artists a basic income and it was a boost for the economy

good.is
6 points·by stared·le mois dernier·4 comments

15 Years of StarCraft II Balance Changes Visualized

github.com
4 points·by stared·le mois dernier·0 comments

Models try to write a Game Boy Advance emulator from scratch

gbaeval.com
3 points·by stared·le mois dernier·0 comments

Games in which you walk (2019)

p.migdal.pl
2 points·by stared·le mois dernier·0 comments

comments

stared
·il y a 5 jours·discuss
Since when Zuckerberg is an AI expert?
stared
·il y a 6 jours·discuss
I am puzzled (and irritated) why there is „rotate left” without „rotate right”. Does any of you know why?
stared
·il y a 10 jours·discuss
There was "It’s Not Enough to Be Right – You Also Have to Be Kind" https://news.ycombinator.com/item?id=21490714

But I think the core part is WHY we want to be right? To prove something to others, or to ourselves? To feel better? As a compulsion? As a gambler's fallacy? Many motivations are less lofty that we dare to admit.

I wasted way to much time arguing online. It was mostly wasted time, and wasted emotions. I mean, I also had many eye-opening and enlightening discussions, but these rarely were fights.
stared
·il y a 11 jours·discuss
Nice!

If you want to play a hyperbolic minesweeper, Hyperrogue features that https://hyperrogue.fandom.com/wiki/Minefield
stared
·il y a 11 jours·discuss
Yes, it gets really hot really fast.

As much as I was tempted to use it on longer projects, I had some reservations about whether it would put too much strain on my MacBook.
stared
·il y a 11 jours·discuss
All experiments with Qwen 3.6 required no more than 48GB Apple Silicon. I believe you can go even further with more aggressive quantizations - one can go down even further.

In any cases, from the economic point of view, running models on laptops make little sense. Even at the pure cost of energy consumption, it might be hard to beat pricing at tokens generated at scale.

At the same time, it is a breaktrough, that will change the game. Previously such vibe coding on consumer device was not hard or costly - it was impossible.
stared
·il y a 16 jours·discuss
I didn't know it has a name. But I have been using similar thing, https://github.com/harbor-framework/terminal-bench-science/p...
stared
·il y a 18 jours·discuss
See also Gribouille: A Grammar of Graphics for Typst, discussed here a week ago https://news.ycombinator.com/item?id=48541062.
stared
·il y a 18 jours·discuss
Violin plots have an interesting reputation (https://xkcd.com/1967/, https://www.reddit.com/r/labrats/comments/91ex4u/is_it_just_..., https://jabde.com/2022/12/22/banned-violin-plots/).

For showing distributions, I much prefer strip plots (https://seaborn.pydata.org/generated/seaborn.stripplot.html), perhaps with opacity, or swarm plots (https://seaborn.pydata.org/generated/seaborn.swarmplot.html) - no averaging with an unknown kernel, no hiding distributions behind a box plot, and the data is directly visible. We also directly see whether it is based on 5, 100, or many more points.

When using histograms, binning is usually more straightforward than kernels. And in any case, the mirror reflection of a histogram is not needed.
stared
·il y a 18 jours·discuss
A year ago, added R to the pipeline (with multiple complications) just to use ggplot2 - even though Python was the main tech.

https://quesma.com/blog/sandboxing-ai-generated-code-why-we-...

Good, that ggplot2 can run inside in WASM, vide https://github.com/QuesmaOrg/webr-ggplot-playground
stared
·il y a 18 jours·discuss
For malware detection, many models are biased for or against detecting a threat (likely a thing that can be adjusted with a prompt).

I suggest tasks cannot be guessed (find, not tell). And 2d charts, both for ROC and pricing, vide https://quesma.com/benchmarks/binaryaudit/
stared
·il y a 19 jours·discuss
> A plain dictionary makes this harder or virtually impossible (you most often will silently fail)

I didn't know about these Typst restrictions. Silent fails are the worst, so if a constructor is necessary to prevent these, good it is there.

Thanks for explaining! (And for developing gribouille in the first place!)
stared
·il y a 19 jours·discuss
My take: does something add value OR is there because were are just used to?

Gribouille is not ggplot2, or other. Syntax is different. Superficial keyword similarity is (usually) a false friend. Reusing a keyword might be useful, but keeping an unnecessary construction is (in my view), a cargo cult.

Typst itself breaks with a lot of LaTeX stuff, and it is good that it does not pretend it is LaTeX-with-Rust, but has a fresh look.
stared
·il y a 19 jours·discuss
Tauri is getting traction in the meantime.

A non-native UI has some issues, but also one clear advantage - it is easier to make a cross-system app with the same looks.
stared
·il y a 22 jours·discuss
Interesting! If I get it right, the API is in the spirit of Observable Plot (https://observablehq.com/plot/), less ggplot2.

In any case, I'm curious whether aes is necessary, or whether it would suffice to drop this function entirely and just use keys in the mapping (similarly for labs). Or, more broadly, whether using patterns from other implementations of the Grammar of Graphics is a conscious decision, or some sort of legacy baggage.
stared
·il y a 22 jours·discuss
It revolves around the sentiment of "go deeper" - but I think it is a double-edged sword. Sure, entropy, tensors and gradients are important - and yes, they are pretty much requirements.

But from what I see, it is the opposite - a lot (if not virtually all) progress in the last decade of deep learning was not because of a fundamental idea, but incremental, experimentally-verified practice. Even though I think there is good intuition for why ReLU is better than sigmoid (tl;dr: last layer is log(sigmoid) ~ ReLU, putting anything different inside kills the gradient), the original paper by Hinton himself was more or less "because it trains 3x faster".

Re-thinking fundamentals might help, but most "let's change the fundamentals" is rarely how it works. Even the most seminal papers, i.e. AlexNet and "Attention Is All You Need", are refinements of existing ideas, and show how they help.

Machine learning is an experimental science. Many mathematically cool ideas do not work. Many engineering ones do.

> I've tweeted before that one of the most important traits in a researcher is healthy paranoia. Be paranoid!

I have seen so many PhDs burned out to cinders; I don't think it is any more a good piece of advice than "depression is good for philosophers". Sure, be a relentless explorer.

> In short, holding on to ideas for too long can actually be counterproductive. Stay open-minded and refuse to let ego cloud your judgement.

Which I think is true.
stared
·il y a 24 jours·discuss
https://xcancel.com/arena/status/2066957802741043641#m
stared
·il y a 24 jours·discuss
Military race isn't a good think either, but you don't want to be on the losing side.
stared
·il y a 24 jours·discuss
If there were European MicroSoft or Google, there would be a preference.
stared
·il y a 24 jours·discuss
Let’s autonomous Russian drones, and Europe is at mercy of two other empires, who capitalize on this opportunity.