gliptic

20 ms·

241.
by gliptic 8y ago
You seriously don't think Zero can beat humans? 6 stone advantage over AG Lee that was trained on human games? If AG Lee behaves like a superstrong human, then Zero handily beating AG Lee is clear evidence that Zero would beat humans.
242.
by gliptic 8y ago
You're right, I misremembered how Master was trained because they already had developed Zero by then. I still don't think there's any reason to believe Zero would do worse against humans than the Lee version. Zero was much st
243.
by gliptic 8y ago
But that isn't true. AlphaGo Zero seemed to be considerably stronger against humans than AlphaGo Lee. Lee managed to win one game against AlphaGo Lee, but AlphaGo Zero won 60 games in a row against top players, as well as all games aga
244.
by gliptic 8y ago
This is actually how it works in Sweden. They check the voter list whether the person already voted during the day and discards the mail-in if so.
245.
by gliptic 8y ago
The 500 ton figure also includes the oxidizer which contain no carbon. The equivalent of a plane's jetfuel, RP-1, in a current Falcon 9 is just 123+32=155 metric tons [1]. [1] http://spaceflight101.com/spacerockets&#x2F
246.
by gliptic 8y ago
There's a much more reasonable explanation for the downvotes. You're making a very dubious claim without even trying to back it up.
247.
by gliptic 8y ago
It also has some funny Easter eggs, like a "Fibonacchos" stand.
248.
by gliptic 8y ago
Gravity doesn't work that way, I'm afraid, as Galileo demonstrated.
249.
by gliptic 8y ago
Well, it was in January of 2015, half a year before the CRS-7 anomaly. Google and Fidelity jointly invested $1 billion on a $10 billion valuation. You can't have been researching this very carefully.
250.
by gliptic 8y ago
As just one example, you say Google most likely saved SpaceX from bankruptcy in 2016. As far as I know, neither Google nor anyone else did any investment into SpaceX in 2016. Where are you getting your information? Also, have you not even c
251.
by gliptic 8y ago
If you knew history, you'd know this has been tried before. The shuttle boosters did this; they barely saved any money after refurbishing. SpaceX tried it too, but had significant problems even re-entering without the rocket ripping ap
252.
by gliptic 8y ago
Why is everyone assuming C is "crash early" as opposed to "undefined behaviour will crash if you're lucky and cause RCE in the worst case, or any weird thing in between". The selling point of the borrow checker (at
253.
by gliptic 8y ago
The refurbishing costs of even their first reused booster was less than half the price of a new booster [1], so something above 50%*70% = 35% saving assuming that the booster is 70% of the cost of the launch. The new block 5 version rocket
254.
by gliptic 8y ago
Yeah, such nuance. Implying that I hold Musk as a "messiah" because I'm defending a backup plan that Musk was asked to continue working on, against people that apparently knew in their heart it was 100% doomed to failure from
255.
by gliptic 8y ago
What exactly are you arguing? You admit it could have been useful, but because they decided to go ahead with another approach because of the circumstances, the effort is useless? Do you also not do backups because, hey, I've never lost
256.
by gliptic 8y ago
Yes, you certainly know better than Dick Stanton, of course https://mobile.twitter.com/elonmusk/status/10166843660831907...
257.
by gliptic 8y ago
Well, you're welcome to try, but it's not like free space bitmaps are unknown to allocator writers.
258.
by gliptic 8y ago
1 bit per byte would require an awful lot of memory access to find a free block. You would need something like a hierarchy of bitmaps to cut down on scanning.
259.
by gliptic 9y ago
It's going to take more than some news article and some heavily criticized studies to establish such an improbable claim.
260.
by gliptic 9y ago
What are you talking about. That gif is showing the fairing separation. There's no sphere.
261.
by gliptic 9y ago
There's nothing false about the sense of security. In fact, after I started using Rust, my C/C++ has become more disciplined as well. The idea that memory safety is "training wheels" is silly.
262.
by gliptic 9y ago
"När du googlar" is literally "when you Google". Ironic, he could translate others but not himself ;). "Datavetare" would be more correctly translated as "computer scientist" and I'm surprised it
263.
by gliptic 9y ago
Well, then you should have noticed what the article was about, which was not to detail a research program about AI safety. Different articles can address different aspects of a problem without being accused of advocating "random action
264.
by gliptic 9y ago
So, in your view, starting MIRI, doing fundamental research into AI safety and advocating for it, is not trying to find the damn light? You exemplify exactly the attitude he's trying to combat. "Oh, nobody knows anything, let&#x27
265.
by gliptic 9y ago
http://www.spacexfm.com/ is a good site.
266.
by gliptic 9y ago
It's only the turbopump that is electric. The efficiency of the turbopump is counteracted by the need to bring batteries along. As I understand it, the electric turbopump makes the plumbing simpler, but it doesn't scale to large e
267.
by gliptic 9y ago
"YC YC YC Rust Rust" is also 100% :). Gotta have some Rust.
268.
by gliptic 9y ago
"Rust Rust Rust Rust Rust" seems to be the optimal number of Rusts with 97.6% success probability. EDIT: This beats it with 99.4%: " Rust Rust Rust Rust Rust "
269.
by gliptic 9y ago
There is LZNA ( http://cbloomrants.blogspot.se/2015/05/05-09-15-oodle-lzna.h... ) that uses vectorization to update and query its statistical models.
270.
by gliptic 9y ago
They describe it like that because it's a long short-term memory recurrent neural network, which is a machine learning algorithm. How exactly were you able to infer that they were lying and were instead using a Markov chain from a few

More ›