317070

33 ms·

391.
by 317070 5y ago
It prevents the artist from signing the same thing twice. So with PGP you might be the owner of a thing, but you might not be the only owner of a thing.
392.
by 317070 5y ago
I once (8 years ago) did a comparison between random, pseudorandom and quasirandom. Quasirandom worked ever so slightly better than the other two in speed and final performance, but not enough to warrant the additional complexity of impleme
393.
by 317070 5y ago
It kind of does, but there are arguably better ways of doing that. You are effectively putting the opponent in a clearly better position, but giving him the task of converting it before any pressure has been built up. That can throw people
394.
by 317070 5y ago
Since Einstein did not observe this extreme situation of a star around a black hole prior to formulating his theory, it does proof him right. It was a falsification observation, and it verified the theory. Apples and goblins should produce
395.
by 317070 5y ago
XOR is not linear in the inputs. I mention this, because Minsky's paper used that fact to show that perceptrons could never work as AI. This pretty much started the first AI winter in 1973. It's an interesting story: https:/
396.
by 317070 5y ago
Because the alternatives are pretty bad (although a middle ground should be possible, like some countries cancel your citizenship automatically when you acquire another): - Being stateless makes it virtually impossible to travel - Most coun
397.
by 317070 5y ago
Because when you hear the other side (pretend to) hang up, you need to not be fooled and also hang up on your side. If you don't hang up on your side, then on older phones you actually have no way of telling the call is still going. On
398.
by 317070 5y ago
I have "research scientist" as a job title. It's a thing: https://deepmind.com/careers/jobs/469515
399.
CIA officials under Trump discussed assassinating Julian Assange(theguardian.com)
9 points by 317070 5y ago|2 comments
400.
by 317070 5y ago
Belgium and Switzerland come to mind? Two highly federalised states.
401.
by 317070 5y ago
National flags and languages are not a 1:1 map. Some flags have multiple languages, and some languages have multiple flags. And that can be "close enough", until you for example serve English speaking people in Ireland the Union J
402.
by 317070 5y ago
But that is not what GPT-3 is. You could say that is what they did _given GPT-2_, but that is underselling the complexity. Some scientist at OpenAI had the hypothesis "what if our algorithm is correct, but all we need is to scale it 3
403.
by 317070 5y ago
For GPT-3, they use the data everyone is using. Their cleaning is unique, but so is everyone's. They don't do anything special, but put in some more effort than most though. As far as I can tell, it is a combination of smart algor
404.
by 317070 5y ago
While Ronald Reagan was in office, he had a colectomy for a colon tumor in 1985 (turned out it was not cancer) [0], and his wife had a mastectomy for breast cancer in 1987 [1]: While Reagan did a lot to try to defund cancer research regardl
405.
by 317070 5y ago
sha256sum is not a continuous function. These functions should be continuous. Usually there are light restrictions on these functions to avoid what you are trying to do. Lyapunov bounded, C2 continuous, ... Control engineers could use a pie
406.
by 317070 5y ago
It does. Monster is also a dutch word, with the same meaning as the English monster (horrid creature), but can also be used as synonym for "a sample". In this case, the name of the village probably comes from the same word as the
407.
by 317070 5y ago
No, and yes. You can build something. But even if you imagine tele-operating it 24/7, there is not going to be something relevant you can do with it around the house. It's not about the programming. And that will not even be takin
408.
by 317070 5y ago
It just looks nice. Once you downsample, you cannot accurately upsample again. If the upsampled image looks real to you, and if you would downsample that again you get the same downsampled image, then you did the best you could.
409.
Highly accurate protein structure prediction with AlphaFold(nature.com)
66 points by 317070 5y ago|6 comments
410.
by 317070 5y ago
I think VAE's haven't been the state of the art since around 2016-2017? They have been squeezed from both directions, autoregressive models on the compression side, GAN's on the generation side. They are still fairly competit
411.
by 317070 5y ago
> Generative adversarial networks, or GANs, are a conceptual advance that allow reinforcement learning problems to be solved automatically. They mark a step toward the longstanding goal of artificial general intelligence while also harne
412.
by 317070 5y ago
It does. I am trading off decentralization, privacy, security, ... with (for Bitcoin) more than how much energy I would use in a year and CO2 I would use in a life, per transaction. They do offer a comparable product, and you need to compar
413.
by 317070 5y ago
You can think of this space as something that tells for every point a path passes through, how long the distance of the path in that point is. So when I connect A and B with a path, and I take the integral of space on that path, I get the l
414.
by 317070 5y ago
Some resource needs to be consumed, but we have payment systems that are currently multiple orders of magnitude more efficient than Bitcoin or Chia, so that is kind of a baseline. (Think Visa or Mastercard) Secondly, Chia is a small coin, i
415.
by 317070 5y ago
> using underutilized existing space, not space from new drives Chia is basically turning wasting electricity into wasting silicium (or whatever tapes/harddrives are made from). Because this "space" needs to be created as
416.
by 317070 5y ago
if people like this, there is an interesting podcast on lyfe and life. [0] Apparently, the people who do research to artificial life, where they create things in silico and try to make life evolve from them, constantly run into a similar pr
417.
From Motor Control to Team Play in Simulated Humanoid Football(youtube.com)
1 points by 317070 5y ago|0 comments
418.
by 317070 5y ago
Not in 2025, except if ITER would have a huge unforeseen failure mode. ITER will have good data in 2035. ITER is mostly a giant experiment in programme management. One could argue that it is the most complex machine built so far by humanity
419.
by 317070 5y ago
Well, yes? All of them will try to show a brothers Wright moment of science history in 2025 (except ITER, which will do more like a sanity check). But building commercial airlines is the step after that and will need more funding and resear
420.
by 317070 5y ago
It is no coincidence. If everything goes to plan, ITER will demonstrate first plasma in 2025, which will be a bit of a PR nightmare for ITER (as it will not have high Q's, and go out of operation for another 10 years after that). So a

More ›