m3at

15 ms·

151.
by m3at 6y ago
I wish so too! I don't expect them to release the code (they rarely do) and they wield their usual "it might have societal impact, let us decide what's good for the world": > We recognize that work involving generativ
152.
by m3at 6y ago
It might not have seen that specific combination, but finding an anthropomorphized radish sure is easier than I thought: type "大根アニメ" in your search engine and you'll find plenty of results
153.
by m3at 6y ago
I wonder why none of the top comments mention Ruffle[1]? It's a wasm flash player emulator, and it provide the necessary sandboxing. That sounds to me like a reasonable alternative to a buggy and now unsupported official player. [1] h
154.
by m3at 6y ago
Yes: https://github.com/ruffle-rs/ruffle
155.
by m3at 6y ago
TVM can do that for you: http://tvm.apache.org/
156.
by m3at 6y ago
To quote the Hayabusa2 team directly ( http://www.hayabusa2.jaxa.jp/en/topics/20201204_ts3/ ) : > You might therefore wonder if we should instead analyse meteorites. However, meteorites consist of the materi
157.
Through-Wall Human Pose Estimation Using Radio Signals (2018)(rfpose.csail.mit.edu)
2 points by m3at 6y ago|0 comments
158.
by m3at 6y ago
Not the GP but I made trading algorithm for a living (not on crypto) and have friends still doing that (on crypto): yes it can work, but in 99% of cases you shouldn't do it for the money. Let me expand: If you can afford to spare part
159.
by m3at 6y ago
You're new so you might wonder why you're downvoted. Hackernews has some guidelines, and while they're far from being perfect they do help to maintain above average online conversation. Please have a read (it's pretty sh
160.
by m3at 6y ago
You raise an important point. The proposed solutions are too many to enumerate, but if I had to pick just one currently I would go for "Rethinking Attention with Performers" [1]. The research into making transformer better for hig
161.
by m3at 6y ago
On the Measure of Intelligence, François Chollet [1] Fellow HNer seems to have liked a lot of ML paper, this is not breaking the trend. This is a great meta paper questioning the goal of the field itself, and proposing ways to formally eva
162.
by m3at 6y ago
This is a great example made out of lego on how counter-intuitive this can of structures can be: https://old.reddit.com/r/lego/comments/g2230l/lego_tensegrit...
163.
Tensegrity(en.wikipedia.org)
4 points by m3at 6y ago|1 comments
164.
by m3at 6y ago
As this is sufficient to cool down the pi even at max load, I assume they considered the esthetic next
165.
by m3at 6y ago
> Sometimes it works, and sometimes it doesn’t, but I always find it to be the least stressful way of addressing disputes. This make it very valuable to me. Resolving disputes can quickly become stressful, especially when as an engineer
166.
by m3at 6y ago
I see a lot of negative reactions in the comments, but at least this part seems fair to me: > That advance will be much clearer once their peer-reviewed paper is published (we should not judge science by press releases), and once the too
167.
by m3at 6y ago
Good advices and a well written article. To add one advice about a common occurrence the author mention: > you might knock it out of the park one year and I might give you something that’s nowhere near as good. Next time, I'm going
168.
by m3at 6y ago
It sounds absurd, but NVIDA has been doing it for a few years, so it certainly can be done...
169.
MinuteEarth: How to Work from Home as a Team(youtube.com)
1 points by m3at 6y ago|0 comments
170.
by m3at 6y ago
Your question is too broad, but to give at least one data point: a data science dev with a few years of experience (1 to 5), can expect between ¥6M and ¥12M/year in a startup. > it should at least be higher than megacorps Yes, with
171.
by m3at 6y ago
The article quote ~17K for N5, do you have an other price reference or was it just a typo?
172.
by m3at 6y ago
Good and concise explanation thank you
173.
by m3at 6y ago
> it would still be trivially easy for criminals to use it Maybe it's not about the criminals? Governments have a tendency to seek to increase control of citizens, surveillance being part of it. Saying it's about "tools fo
174.
by m3at 6y ago
If you want to read the source, I recommend skipping to their follow-up paper (same authors): https://arxiv.org/abs/2009.07118 Edit: tangentially related but for those who like to have a glance from their phone, arxiv-
175.
by m3at 6y ago
Thanks, those are good charts
176.
by m3at 6y ago
For more of this, please look at: http://www.tylervigen.com/spurious-correlations
177.
by m3at 6y ago
Atlas Shrugged, by Ayn Rand
178.
by m3at 6y ago
It's possible Google will keep paying what is peanuts to them not to be technically an absolute monopoly (by keeping Firefox alive at insignificant usage level)
179.
Phoshpine Spotted in Venus’ Atmosphere(arstechnica.com)
1 points by m3at 6y ago|0 comments
180.
Cellular Automata as a Language for Reasoning(kaggle.com)
2 points by m3at 6y ago|0 comments

More ›