x3tm

10 ms·

31.
by x3tm 8y ago
>Ok. I brought it up because "I don't see race" is a popular rebuttal to any attempt to discuss racism Yes, I got the good intention behind your message. No worries. > I think there's a lot of valuable ground to di
32.
by x3tm 8y ago
Thanks for your clear reply. There's only one thing though from what you mentioned that is not vague or subjective: > It is the one place where AI can have a really positive impact on the world. Can you develop on this please? Why d
33.
by x3tm 8y ago
My comment was that I dont see race _when it comes to science_. Science is not influenced by race. Race is not relevant for scientific discussions. I was criticizing the fact that the mixing between race and science is bad at best. Wherever
34.
by x3tm 8y ago
Strange article. Why give Watson, who most people forgot about anyway, publicity in this way? Seems to me that this kind of articles fuels racism by introducing bigots to a great bigot they can look up to, more than it actually fights racis
35.
by x3tm 8y ago
How about universities?
36.
by x3tm 8y ago
Science should be above this. Inspired by your comment, I'll make my point more graphically: Nazi camps were also suitable for medical and spychological testing. Nazi and even interned doctors did all sorts of experiments on prisoners.
37.
by x3tm 8y ago
Merkel is a physicist. She's great. But then again many physicists were also convinced Nazi officers.
38.
by x3tm 8y ago
Or how he shutdown the drones gaming unit, and his constant efforts towards peace in the middle east by refusing to destabilize countries, etc. Great man.
39.
by x3tm 8y ago
What I struggle to understand is how could esteemed AI researchers like LeCun and many others work for such a company to give it even more powers and edge? How can they justify this as scientists? The huge influx of money into research (her
40.
by x3tm 8y ago
Or a significant investment in dongles.
41.
by x3tm 8y ago
I suppose a good example is whatever runs in NASA's control room or (to a lesser extent?) in control towers.
42.
by x3tm 8y ago
I don't think it's his biggest contribution to physics; how about QED? The path integral formalism or the diagrams, albeit being mathematically obscure at first where quite clear from the intuitive viewpoint once he explained them
43.
by x3tm 8y ago
Gmail seems to have 3 nines, although I couldn't find a better reference than this [1], where other services are included: > [Google's infrastructure] delivers Gmail and other services to hundreds of millions of users with 99.9
44.
by x3tm 8y ago
Exactly. Effective communication is what made Feynman a legend. Sadly, many extremely smart and profound scientists are quite incapable of (or not interested in) conveying clearly their thoughts to general audiences.
45.
by x3tm 8y ago
What are the main factors that could lead to such a situation?
46.
by x3tm 8y ago
Better/faster communication can facilitate work, but it's not the same as causing a breakthrough like say new data/theory. In that regard, English as a unified scientific language has probably a bigger impact than the interne
47.
by x3tm 8y ago
I am studying your book on DL right now. I came here to have a short break ... and there you are :-) Thanks for the book. It's great!
48.
by x3tm 8y ago
Impact on society doesn't mean impact on science. What scientific breakthroughs were due to the internet? Yet, the internet had/has an unprecedented impact on history. On the other hand, when someone starts having public opinions
49.
by x3tm 8y ago
A key point of that piece in the Atlantic, if I remember correctly, was that there are no breakthroughs now because we already know, for the most part, what is there to be known. We know the most important ideas/theories about the worl
50.
by x3tm 8y ago
In other words, he overfit his model (T).
51.
by x3tm 8y ago
Interesting observation!
52.
by x3tm 8y ago
Or planes can fly, right? I understand this, however it's not a fair point. Without proper understanding of 'swimming' and 'flying' these technologies wouldn't be possible. So the link between thinking and AI i
53.
by x3tm 8y ago
I think if we push this to the limit we would all be writing assembler or making circuits :)
54.
by x3tm 8y ago
Age is a pretty good argument actually for computer languages IMO. More abstraction is better, and this usually comes with time and progress. Otherwise we would still be using Assembly, which incidentally should be faster than Fortran sinc
55.
by x3tm 8y ago
We commonly used the GSL library. http://www.gnu.org/software/gsl/
56.
by x3tm 8y ago
Exactly. That's why I say that the speed argument is a bit obsolete by now. We can have the best of the two worlds with python! Elegance/simplicity and speed. The focus on the speed of execution also downplays two other important
57.
by x3tm 8y ago
Thanks. I admit what made me chose TF is the support from Google which guarantees somehow that the tool will stick around for some time. And the numbers of contributors to the library. Where do you see PyTorch in the near future?
58.
by x3tm 8y ago
Which are not free for the most part, right? As of 2019, the speed argument is mostly academic though. IMO. Few people do cosmo simulations on a daily basis to really care about this.
59.
by x3tm 8y ago
They use the Fortran 90 version I suppose. I should've said that I'm a theorist though and HPC is not really an issue for us (And if I do want something to be fast I'd use C.)
60.
by x3tm 8y ago
Yep. That's the community I had in mind. Particle physicists love Fortran :-) However, I'm not so sure about the speed argument anymore. And in any case for most applications (which are rarely big simulations) it's not worth

More ›