rspeer

51 ms·

421.
by rspeer 9y ago
That advance happened by 1992, with TD-Gammon [1]. Our hardware and software are clearly much better now, but this seems like a solid example of what the GP said: the conceptual framework has stayed the same for 25 years. [1] https:&#x2F
422.
by rspeer 9y ago
AI is just a tool, a way to automate making more kinds of decisions. The decisions are not better or inherently more right by virtue of taking a human out of the loop. In fact, they're usually worse, because humans are intelligent
423.
by rspeer 9y ago
Enough of this false equivalence. You can call what I do "introducing my own biases". I am "biasing" NLP algorithms toward treating people equally whenever possible, toward not discriminating based on their name or how t
424.
by rspeer 9y ago
"Non-profit" is a whole different kettle of fish than "non-commercial". It doesn't mean you're not selling anything. It doesn't mean you're morally good. It just means you registered for a particular
425.
by rspeer 9y ago
Is it actually shocking? Isn't it just a statement the Orthodox church would agree with ("there is no god except God") written in Arabic?
426.
by rspeer 9y ago
I consider it very important to fight bias in AI, but I fundamentally do not understand this article's take. The thing stopping AI researchers from obtaining unbiased training data is not that we're waiting for laywers to give us
427.
by rspeer 9y ago
And what if an open source project wants to enter this field?
428.
by rspeer 9y ago
The criticisms of the Implicit Association Test don't apply to testing AI systems for racial bias. You ask a word2vec-based system about predominantly-black names and it associates them with negative emotions, gives them negative senti
429.
by rspeer 9y ago
And also cannot be used in open-source projects. CC-BY-NC amounts to saying: you can play around with this in demos and academic projects that no lawyer would ever go after anyway, but you can't use it for real.
430.
by rspeer 9y ago
Not that I have experience in the matter, but I have heard that spending bolivar in Venezuela means waiting in long lines for goods with artificial prices, whereas you can buy whatever you want in USD. I suspect that the bolivar/BTC tr
431.
by rspeer 9y ago
I don't care about a bunch of Python 2 research-quality code. The amazing part of the project, to me, is all the data that they make available in one place.
432.
by rspeer 9y ago
That's a good point. It seems about right that license terms are all about the cases where the code is used "commercially", for some sense of "commercial". If the code is never used commercially, there's never
433.
by rspeer 9y ago
Someone explain this part to me: > I would be happy if Zcash and Bitcoin could serve as a gateway from an unstable currency (Venezuelan Bolivar) to a stable currency (EURO/USD) How would this actually happen? Who is on the other sid
434.
by rspeer 9y ago
My understanding is that you would have to believe that they were kidnapped or otherwise under duress at the time that they did the trusted setup, which was on video. The trusted information (supposedly) can't be given up now because
435.
by rspeer 9y ago
You can't release your code in a way that it can be used as a dependency, except by requiring your downstream users to use the License Zero payment platform. That's an encumbrance. You also can't distribute it along with GPL
436.
by rspeer 9y ago
CC-NC is not a license for code. Creative Commons has spent a while trying to pin down what counts as "commercial use" for art and data, but they'll tell you flat out that code is out of scope. Ideas like "dependencies&q
437.
by rspeer 9y ago
You're not describing a strict dependency. You can just buy Tableau and use it.
438.
by rspeer 9y ago
Setting aside the practical detail that not everything is node.js: Nobody would want to sell commercial software under the terms of "you have to license our software A, but it won't work unless you also license software B from som
439.
by rspeer 9y ago
You may find it hard to believe, but the only reason these chains were ever viable businesses is that middle-class people would go there, by choice. Yes, they got worse. Sticking with the example of Friday's, it used to be a reasonable
440.
by rspeer 9y ago
You probably have to bring the trash with you for a while, so ending up with a smaller empty bottle can be preferable. There aren't a lot of public trash cans. (Before I realized this, I was once perplexed by where to throw out a food
441.
by rspeer 9y ago
I doubt more people are cooking, it's just that the largest chain restaurants have gradually let themselves go to shit and left themselves in a position with no demand. There are more options than ever for good, casual food. And there
442.
by rspeer 9y ago
Ah. I hadn't caught the part where it relies on Intel's Attestation Service. That is indeed weird.
443.
by rspeer 9y ago
If I'm understanding you, this seems like unwarranted cynicism, and I say this as someone who is quite cynical of blockchain technologies. Would you also criticize the current state of Ethereum as "proof of Nvidia" and imply
444.
by rspeer 9y ago
There's still that "Firefox is installing updates" modal that shows up the first time you launch the new version. But after that bad first impression, launching Firefox 57 is very snappy.
445.
by rspeer 9y ago
I don't think you really addressed my example. I'm talking about a difficult case of distinguishing colors, where you don't get to look at them next to each other, and hypothesizing that it would help if the colors were descr
446.
by rspeer 9y ago
In my experience with local politics, "affordable housing" can mean any of three things: - Subsidized housing - Public housing - Increasing density to lower the actual price of housing But homeowners (overrepresented among voters)
447.
by rspeer 9y ago
So I'm fully aware that people have experimented with the Sapir-Whorf Hypothesis for decades, and if it were true it would have been conclusively demonstrated by now. But isn't color names one of the last refuges of Sapir-Whorf, o
448.
by rspeer 9y ago
I was responding to the claim that "git doesn't treat large files any differently than small ones". I was not saying anything about dat or git-annex. Now it's clear that you are treating your large files differently fr
449.
by rspeer 9y ago
GitHub definitely treats large files differently from small files (it rejects them), and a Git repository you can't push to GitHub is very different from a typical Git repository. Other than that, Git treating large files the same as s
450.
by rspeer 9y ago
Git is not good for large files. If the suggested solution is to use git-annex, I will say that git-annex is so poor in usability that the majority response to "you can get the data via git-annex" is "oh well, I'll try s

More ›