argonaut

91 ms·

931.
by argonaut 10y ago
Agreed, but most people won't understand stories until/unless they have a bunch of friends on Snapchat, so it's not the best aspect to start with.
932.
by argonaut 10y ago
People need to stop focusing their criticism on how stupid the app is or how they're going to lose all their users. It's almost a meme by now of how people say this every time Snapchat is brought up. By now it's a foregone co
933.
by argonaut 10y ago
> Brookings and the American Enterprise Institute claim to have hatched a bipartisan consensus plan for reducing poverty. As exciting as that sounds, the details of the plan, unfortunately, won't be available until David Brooks unve
934.
by argonaut 10y ago
Wait what? You have it the other way around. Rail will still be dramatically more efficient and cheaper than electric autonomous cars. And inter-city subways will still be more efficient and cheaper than electric autonomous cars. They have
935.
by argonaut 10y ago
Opposing the construction of more housing will just maintain/accelerate the rate at which residents of SF will be driven out.
936.
by argonaut 10y ago
Which is a perfectly accurate description of many architectural improvements in deep learning (e.g. residual networks).
937.
by argonaut 10y ago
Credit Suisse just copied the numbers from Global Trade Alert. Did you even look at the reports on the Global Trade Alert website? Or did you just read the "About" page? Mentioning Trump is just FUD and indirection. You can mentio
938.
by argonaut 10y ago
> increasing market-rate housing alone may not make an appreciable difference in prices where it matters most from a macroeconomic perspective: for low-to-mid income residents. This may well be possible to argue, but you're going to
939.
by argonaut 10y ago
But what's the solution to this problem? Ban wealthy people from financing lawsuits? That's never going to happen. What about ban people with malicious intent from financing the lawsuits? You'll rarely be able to prove the in
940.
by argonaut 10y ago
Irrelevant to my point?
941.
by argonaut 10y ago
> Imagine the amount of damage something like NAFTA has done to the american working class - multiply that with 200 years. Citation needed. This is still a controversial issue in policy, and anyone certain of any one side has no idea wha
942.
by argonaut 10y ago
Your links cites a report that itself gets its numbers from the "Global Trade Alert," which is an org that aggregates a variety of trade-related actions published by the governments of each country. Only a fraction of these are a
943.
by argonaut 10y ago
PAQ does not use autoencoders, that's the difference. Neural network models for compression are extremely slow and can use a crapton of memory (one top benchmark result uses 32GB of RAM): https://www.quora.com/What-is-t
944.
by argonaut 10y ago
The article does not evaluate the quality of the embeddings, so there's not much to say about all of that.
945.
by argonaut 10y ago
Some of the other movies (the really bad examples) are out of sample. The last example is also from training set, though.
946.
by argonaut 10y ago
To add on to this, it's also a pure tech services play, which IMO is generally a bit harder to monetize than a product company, but generally a lot easier to sell in an acquisition. Which doesn't make this sitatuion any better.
947.
by argonaut 10y ago
Correct me if I haven't looked into this closely, but one glaring problem is that all the results are from the training set. So it's not surprising you get something movie-ish by running the network over a movie it was trained on
948.
by argonaut 10y ago
Dropout helps with convergence/optimization, sure. The existence of a global minimum says nothing about the time required to reach it. Important to note that dropout isn't as common anymore; it's not a huge win.
949.
by argonaut 10y ago
A quadratic function is a convex function with a global minimum. Doesn't mean it's a good model for much.
950.
by argonaut 10y ago
No it doesn't have to. Stop moving the goalposts. The US has a free press, and under no definition does free necessarily require fair. It would be better if it had a fair press, but it is not necessary for freedom. You're again
951.
by argonaut 10y ago
DeepMind is probably losing Google tens of millions a year. Now, it may pay off financially in the future, but it's hard to disentagle its effects given that Google already had a deep learning research group, Google Brain.
952.
by argonaut 10y ago
For context, if true (the paper has not been peer reviewed yet), this confirms what has already been suspected and empirically evaluated about deep learning: that the non-convexity of neural networks is not an issue.
953.
by argonaut 10y ago
The US media as a whole may not be asking critical questions, but that's not the point of free press. The fact is that the US media is free to ask critical questions, and therefore some , though not all, of the media ask critical ques
954.
by argonaut 10y ago
Not only does this have nothing to do with what I said, but it's also ridiculous even if it had any bearing on what I said. If Julie actually won, Github would have agreed she was sexually harassed. They did not do so, and she was not
955.
by argonaut 10y ago
But they can be challenged. That's why you're reading an article about it. If you have a judge that is biased, it is probably harder to challenge his sentences than if you had an algorithm that you proved was biased.
956.
by argonaut 10y ago
The article ( http://www.dailydot.com/business/julie-ann-horvath-names-git... ) says: "complicit in the actions of both Tom and Theresa Preston-Werner and even admitted to plotting with Theresa Preston-Werner to ge
957.
by argonaut 10y ago
pron makes no claim that neural nets are on the extreme end of ad hoc.
958.
by argonaut 10y ago
That's just the public report. It is perfectly consistent for a company to publicly say nothing wrong occurred, while quietly firing some of the few individuals involved. Personally, the sudden-ness and lack of reasons for firing lead
959.
by argonaut 10y ago
His writeup: https://zachholman.com/talk/firing-people We cannot know for sure whether it has anything to do with his firing (and he doesn't mention it in the writeup), but Julie Ann Horvath did accuse Zach of har
960.
by argonaut 10y ago
Completely agreed with pron's comment. Also, the process by which they designed the new architecture is adhoc. They saw a problem, and then tried a bunch of solutions, and got something to work. Sure, their effort was guided by intuiti

More ›