sgillen

37 ms·

1. 421.▲
   by sgillen 8y ago

   Thanks for your response, I had actually heard about Yuri before and was hoping you would respond to this post! I have a lot of questions for you if you don't mind, maybe you don't want to answer any/all of them that's f
2. 422.▲
   by sgillen 8y ago

   I agree with the other guy, there is way more interesting stuff to read than there is time to read it. I find it helps to really have a goal when setting out to learn new stuff. Rather than just "I want to learn more about blah" T
3. 423.▲
   by sgillen 8y ago

   Yes I agree, I think the idea here is to provide applied machine learning as a service to game devs.
4. 424.▲
   by sgillen 8y ago

   Pytorch's official one is very good: https://pytorch.org/tutorials/beginner/pytorch_with_examples... They even start with NumPy, as the article does.
5. 425.▲
   by sgillen 8y ago

   I think they are using the wrong word here, I think by "determinism" they mean interpretable. You train your big neural network to classify images, it works fine on the test set but then one day in the wild you discover your netwo
6. 426.▲
   by sgillen 8y ago

   I guess the point here is that in this article we implement some of the magic that the higher level frameworks use, which I think will be useful later when writing and debugging our high level code. But I see your point which is that it wou
7. 427.▲
   by sgillen 8y ago

   That would be a great contribution but unfortunately I think not an easy one to provide. A few issues 1. Most BLAS libraries aren't deterministic. Floating point numbers aren't in general associative, (a + b) + c =/= a + (b
8. 428.▲
   by sgillen 8y ago

   I still think it's a useful exercise for people like me who learn best by implementing a thing. Tensorflow, Pytorch, etc introduce a lot of magic for someone who is completely unfamiliar with the area.
9. 429.▲
   Ask HN: Game devs, what is the hardest part about designing an AI for your game?

   89 points by sgillen 8y ago|83 comments
10. 430.▲
    by sgillen 8y ago

    Possibly California's legislature wants to avoid fractured governance of the internet in the US if they still have hope in net neutrality nation wide? Possibly if some of those grim predictions start to come true they will take action.
11. 431.▲
    by sgillen 8y ago

    That's a real shame, a (presumably) well intentioned attempt to increase access to educational resources winds up restricting access instead. I wonder what the motivations and intended outcome of the lawsuit was.
12. 432.▲
    by sgillen 8y ago

    I think the key word when talking about status wrt to research is “prestige”. Just about everyone submitting to Nature I’m sure legitimately wants to share their knowledge and get feedback. But I’m sure they’re also really hoping they get t
13. 433.▲
    by sgillen 8y ago

    And to add on, octave is a good open source replacement for base matlab with near identical syntax.
14. 434.▲
    by sgillen 8y ago

    To add to this: Simulink and select other matlab toolboxes (Which btw cost extra). I think base matlab can be supplanted very well with scientific python these days.
15. 435.▲
    by sgillen 8y ago

    I think I mostly agree with the authors premise that "We are all status seeking monkeys". It's just that the type of status people seek is different across populations: > Some people find status games distasteful. Despite
16. 436.▲
    by sgillen 8y ago

    Is it “incredibly” high though? I think you have legitimate criticisms but compared to other news aggregator/ social media sites I think the quality is very high.
17. 437.▲
    by sgillen 8y ago

    You can still do cutting edge work with 3 small layers, for example most of the advances in robotic control (using neural networks) use networks with similar sizes.
18. 438.▲
    by sgillen 8y ago

    I think determining the wind speed and direction 36 hours in advance is the tricky bit.
19. 439.▲
    by sgillen 8y ago

    I see the appeal. Sometimes you want a small tablet sized screen but you want to carry around and use something with a phone sized form factor.
20. 440.▲
    by sgillen 8y ago

    Is Alienware a good choice for laptops now? I’d always heard they were overpriced for the hardware you get.
21. 441.▲
    by sgillen 8y ago

    That's the trade off I guess, because of the JIT you have a slow initialization followed by very fast code. That's not really good for a lot of workflows though (if you're running a smallish script over and over for example).
22. 442.▲
    by sgillen 8y ago

    Not convinced it will completely replace current techniques, if it does start significantly outperforming more old school control systems we’ll still need a lot of progress in explaining what the controller learned and proving that it will
23. 443.▲
    by sgillen 8y ago

    All the new optimization code has C++ and/or python (2.7) bindings. I don’t think TRI has made ROS wrappers but it could conceivably be done without much pain I think. Not sure there is much in there about new stuff from BD. There is s
24. 444.▲
    Ask HN: Best textbooks for learning operating systems

    47 points by sgillen 8y ago|12 comments
25. 445.▲
    by sgillen 8y ago

    Do you have any good/favorite textbooks (or other resources) to learn those from?
26. 446.▲
    by sgillen 8y ago

    As to your last point those variables could be anything, even in the simple examples you list. Size could be a tuple with the shape of some multidimensional thing, it could the the string "Large" temperature could also reasonably
27. 447.▲
    by sgillen 8y ago

    Why emacs though?
28. 448.▲
    by sgillen 8y ago

    Sounds like you almost just want a regular "dumb" watch.
29. 449.▲
    by sgillen 8y ago

    I’m not sure what you mean, what about python is notoriously unfunctional?
30. 450.▲
    by sgillen 8y ago

    it depends, I think some people do advent of code did it as sort of race every day, in which case I think C++ would be a pretty big challenge IMO.