sgillen

33 ms·

331.
by sgillen 7y ago
This was going to be my advice. But honestly if you can find a position you are more of a natural in you will feel a lot better about anything without changing anything else.
332.
by sgillen 7y ago
emacs&
333.
by sgillen 7y ago
Part of the problem is the hype in the other direction. I guess there are those who claim that this sort of result constitutes super human intelligence or something close to AGI. When in reality this is a hyper specialized system that has t
334.
by sgillen 7y ago
I see a lot of comments downplaying this achievement, saying it's not impressive or it won't be impressive until X condition is met. I welcome skepticism and criticism for this sort of thing, and think most of it that I've se
335.
by sgillen 7y ago
Yeah I agree with non-entity. I recommend maybe finding some projects you're interested in USING, creating a new project, and then hopefully contribute any bugfixes you can manage to find. I find just opening up a good bug report is a
336.
by sgillen 7y ago
Thanks, I wasn't aware of studies in this area. Do you know of studies that compare breaking traffic laws of cyclists in different countries?
337.
by sgillen 7y ago
You're being down-voted for the tone of your post I think. But I think it's an interesting point. If there is evidence that US bikers have a culture that does not following traffic laws that can and should seriously impact policy
338.
by sgillen 7y ago
can you have to be more specific with what scenarios you are facing where you need this? If ssh and rdp are unavailable what is available?
339.
by sgillen 7y ago
Personally I think it would be useful to focus on the "easily build physically accurate simulation environments for robots" using blender part. IMO it makes the most sense to try and make this created environment into an OpenAI gy
340.
by sgillen 7y ago
To be fair I think Matlab is still heavily used in industry, which I view as a direct result of being so dominant among students and researchers. Maybe not for machine learning, but for controls engineering, signal processing, stuff like th
341.
by sgillen 7y ago
How much cooperation is needed though? It seems to me that all that numpy pandas etc. need to do is maintain a stable API, which they already do AFAIK.
342.
by sgillen 7y ago
I've had a hard time learning lisp. I feel like I don't really know what's it's good at, or what to do with it. To remedy that I've tried learning elisp to do some fun stuff with emacs, but usually I end up fighting
343.
by sgillen 7y ago
It means admitting that in some situations strongly typed is a better way to write code.
344.
by sgillen 7y ago
I've never seen it before, to be honest though I don't see this being a successful model. The people not playing every day are already those who are not really hooked on your game, I think most people that come back to a game afte
345.
by sgillen 7y ago
I certainly like text manipulation (well string manipulation at least) in python much more than other languages I use (C, C++, Matlab). It might just be the case that python is the best tool in the authors tool belt for this sort of work. I
346.
by sgillen 7y ago
Quantum computers definitely exist, as to when they will be useful is anyone's guess. Furthermore my money is on a nation state (probably US or China) being the first to get a useful quantum computer. I expect that will be as classifie
347.
by sgillen 7y ago
Well I already have the machine and the electricity is free. That's the only reason though, haven't needed to mess around with AWS much.
348.
by sgillen 7y ago
I use ssh. depending on what I'm doing I'll sometimes develop lcoally, then push changes and run experiments on the big machine. Sometimes I want to explore but still have the big guns at the ready, I use jupyter for this and it w
349.
by sgillen 7y ago
Not sure if this works for you but I'm a big fan of having a very powerful desktop that I remote into from a cheap laptop. Especially if you're serious about AI (or gaming, if you can live with not gaming on the go). I do this wit
350.
by sgillen 7y ago
They do at least address this: > Below the top 10, some items of note include Arduino at No. 11 and HTML/CSS at No. 12. In previous years, some readers have complained that neither should appear on a list of programming languages. I
351.
by sgillen 7y ago
To be fair IEEE spectrum does cater to exactly the people most likely to be in those niche areas.
352.
by sgillen 7y ago
I think most people don't fit squarely in any of the camps. Most people are making compromises between virtues of all the categories, they might just weight them a bit differently.
353.
by sgillen 7y ago
I think BLAS is a spec rather than one library. So your version may or may not be fortran. I do think the original "reference implementation" was written in fortran, which is sometimes called "The BLAS library" but I thi
354.
by sgillen 7y ago
>> Is rust suitable for data science workflows? >> Right now I have to say that the answer is “not yet”. I’ll definitely reach for rust in the future when I need to write optimized low-level code with minimal dependencies. Howe
355.
An Introduction to Generic Programming(generic-programming.org)
2 points by sgillen 7y ago|0 comments
356.
by sgillen 7y ago
Not sure I agree with you on SNOPT. There are many many tools out there that I see people use for convex (and non convex) optimization. Especially considering a single license costs $6000 I would be surprised if SNOPT even captured a plural
357.
by sgillen 7y ago
C++. I know it, I use it, I don't usually like it. A lot of the language just feels like a hack. I usually much prefer to write performance critical code in C but sometimes the utility of C++ or of certain C++ libraries outweighs that
358.
by sgillen 7y ago
I do actually agree that we should apply more funding to replicating results. I disagree that a video log is an efficient way to ensure replicatability. In many disciplines this is not useful at all (math, high energy physics). And in most
359.
by sgillen 7y ago
I totally agree with the transparency thing. I would love to see something like open review ( https://openreview.net/forum?id=ryQu7f-RZ ) become more common, and even have something like this replace the traditional peer revi
360.
by sgillen 7y ago
There are already many companies (including all the ones you listed) that would love nothing more to monopolize / take over the industry. There really are not that many companies that can match the cult like following that Apple or Tes

More ›