svantana

81 ms·

421.
by svantana 8y ago
I agree, but two major issues here: - licensing will be difficult without a paywall, I'm pretty sure this site is "illegal" - there is no way to keep tweaking manually to correct bad sounding parts etc, which limits usefulnes
422.
by svantana 8y ago
I agree, but this was about free/complimentary wifi services, so in these cases you are not really buying anything. I think most of them are also blocking porn sites, etc.
423.
by svantana 8y ago
> perfect execution is rare and almost guarantees success Isn't that a tautology? Is there any reliable measure of execution other than the success that follows?
424.
by svantana 8y ago
Yes spherical emission wouldn't work. Like almost all long distance communication, they use parabolic reflectors.
425.
by svantana 8y ago
Well of course -- any machine/language is a map from inputs (programs) to outputs, and the pigeon hole principle enforces that if some strings are shortened, others must be lengthened.
426.
by svantana 8y ago
> Certain fields will need to undergo a process of radical deregulation in order to give the minority of minds that are capable of our greatest feats of creation the leeway to experiment and to play, as they deliver us the wonders on whi
427.
by svantana 8y ago
Not necessarily, they could run them at a higher clock and perhaps put two or more of them on the motherboard for the high end units. Keep in mind that the iphone 8 is already on par with their fastest 13-inch macbook pro.
428.
by svantana 8y ago
First off, that sounds like baseless speculation. Second, there might be a causal relation between lacking motivation/interests and being jobless. It's not a representative sample of the population at large.
429.
by svantana 8y ago
Thanks! I find it weird that performance-focused javascript projects don't have live benchmarks up so you can judge for yourself. Personally I've found that a semi-clever matrix multiplier in c++ compiled with emscripten can give
430.
by svantana 8y ago
Speculative ad hominem aside, the point in this case was to learn to write code so that good compilers can output performant SIMD-heavy code. It's not trivial and can include a lot of trial and error and looking at assembly. But once y
431.
by svantana 8y ago
> It's the huge matrices where performance matters I wouldn't say that, just look at CGI where a lot of effort is spent making fast 4x4 matrix multiplies. But yeah, a general matrix mult needs to scale, but a lot of the time
432.
by svantana 8y ago
This looks like it's only for 3x3 and 4x4 matrices, e.g. what you usually use in 3d graphics. Or is it possible to do arbitrary matrix multiplication with this lib?
433.
by svantana 8y ago
I'm a big fan of writing simple, portable code that is structured enough to let the compiler optimize it to perfection. If you do it right, a naive A' * B multiplier can be within 20% of BLAS performance, at least for non-huge mat
434.
by svantana 8y ago
Performance is highly dependent on task and graphics is extra tricky to compare, but looking at geekbench, it seems the Samsung S9 is about halfway between iPhone 7 and 8/X. So half a generation behind is more accurate.
435.
by svantana 8y ago
Dawnbench [1] is such an effort (you will need to work out the petaflops yourself from time x performance, but it lists cloud computing cost which probably is more relevant), and MLPerf is an upcoming one [2]. [1] https://dawn.cs
436.
by svantana 8y ago
>Cui estimated that, given the water-based battery’s expected lifespan, it would cost a penny to store enough electricity to power a 100-watt lightbulb for twelve hours. Oh dear, I hate university news "journalism" with a passi
437.
by svantana 8y ago
Nevermind those years, it's already here: https://blogs.msdn.microsoft.com/directx/2018/03/19/announci... From the looks of it, the major engines are on route to supporting it, remains to be seen th
438.
by svantana 9y ago
Norvig's title is "Director of Research", presumably this is a different position than "AI Chief". Source: http://www.norvig.com/
439.
by svantana 9y ago
You mean that there exists at least one hash such that hashFunc(paragraph+hash) = hash? Yup, that probability converges to 1-exp(-1) = 63%. So it's no guarantee, but if you have some wiggle room in the paragraph (change the wording sl
440.
by svantana 9y ago
Well arguably facebook the product has been the problem, not facebook the coorporation . Those other products you mention don't have the same "put all information about yourself on here" design, and nowhere near the amount
441.
by svantana 9y ago
Well, in the same spirit as NSynth, but miles ahead, we have: https://heartofnoise.com/products/galaxynth - morph all types of sounds, laid out on a 2d canvas (made by me) https://soniccharge.com/synpla
442.
by svantana 9y ago
As a researcher and developer in the audio field, it's disheartening that the only things to get crossover attention is stuff from google with some buzzwords attached. There is so much amazing stuff being done, but it's hard to su
443.
by svantana 9y ago
That makes no sense, high earners are the least sensitive to (fixed) costs and are more likely to choose location based on attractiveness. Who would move in if a high earner moved out? Supply and demand have a way of taking care of these
444.
by svantana 9y ago
I wasn't saying it's a good idea, just pointing out that it would be the optimal daylight savings. But I don't think you'd notice it too much because you're just shifting the clock ~2 minutes per day.
445.
by svantana 9y ago
> without much practical thought as to increasing controllability That's not fair, Waymo and others are using redundant backup systems and obviously try to think of every failure case they can. Besides, aren't the remote-contro
446.
by svantana 9y ago
The amount of daylight is not changed by setting clocks. The point is to shift daylight hours away from "sleeping time", but still not having to get up when it's dark. The ideal daylight savings would be to synchronize clocks
447.
by svantana 9y ago
> Raise taxes on them enough and they move to NJ or CT and it's actually a net loss for the city. Theoretically, increased taxes shouldn't affect the cost of living, because the supply is fixed, and the market price is way abov
448.
by svantana 9y ago
The density of NY is the cause of the high costs of living, making it even denser is unlikely to bring costs down long term. It would make more sense just to accept that living centrally is a luxury not everyone can afford, and instead chan
449.
by svantana 9y ago
Plenty of places offer cheaper labor than China. It's more about the quality of their work, although relatively cheap labor is obviously part of the equation.
450.
by svantana 9y ago
Can someone explain what the downside is? I'd much rather the rich spend their money on art than activities with huge negative externalities like flying around in private jets, buying land or employing servants.

More ›