Remember? The one where computers performed at par with humans on a very closely related task to this?
Like, I understand why you're aroused by this impossible-to-overcome flaw you're imagining, where computers could never guess what the "it" is referring to - but it turns out they don't work like the text parsers in 80s chat-bots, and are capable of correctly interpreting these references. They continue to improve at exactly this kind of work.
And hey, do you know what the "deep" in "deep neural net" means? It means there's more layers of nodes between input and output. It's not something metaphysical or subjective - if you look at the diagram of how nodes are connected you can describe how deep it is. Do you object when someone calls a swimming pool "deep"? That's pretty much the same mistake you're making.
And is this thing like a brain? Well, it can do some of the same tasks, and some of them better. Does it use the same mechanisms to do that work? The answer is somewhere between "obviously not" and "who cares?". Planes don't fly the same way birds do - but they were inspired by birds and we call the jutty-out parts "wings". Neural nets were inspired by our (limited) understanding of neurons - and it turns out that (just like wings) they don't have to work just the same way in order to create useful results. How would you respond to someone on an aviation board saying "Those things don't work anything like real wings, they're nothing like birds"? Do you see how, while true, that's kind of not an interesting point to make on every article?
In the end, the key question about a plane is whether it flies - whether it does the thing it was made to do. And neural nets are getting better at flying all the time.
I think you're right that "car ownership" would become less accessible over time, but I think "transportation by car" is likely to get very cheap. I think we'll see ubiquitous "Driverless Uber" style services pop up everywhere (in a variety of flavors), and that's how the non-wealthy get around (at least in urban settings).
But I also think the transition will take a while, especially in rural areas - enough so that I don't think the transition will be too brutal for most people economically. If anything, some may benefit from cheaper human driven cars as they're essentially discarded. But yeah, further out in time it will definitely not be good for people who like driving - just as it's currently very expensive/awkward to maintain and use a horse drawn carriage.
I think some people are jumping the gun a bit right now, but once driverless cars are reliable they're going to go from niche to everywhere very quickly.
I never open the Youtube app, and only end up on Youtube through a link.
And I never want to open random links to Youtube (or Twitter or Facebook - or any of the preloaded apps I can't remove) with some dumb special app - the browser version is fine, doesn't force me to log in, and stays in its tab where I put it.
Hopefully Google gets in some fight with everyone else (and, uh, itself) so that all this garbage stays in browser tabs where it belongs.
I think a lot of people are curious about blockchain technology because Bitcoin has been in the news - people don't want to miss out on the new thing in technology. It's natural that IBM and Oracle are going to use a wedge like that to try to get in the door with people, but I don't think that many businesses are going to have problems where blockchain tech is an important part of the solution.
I'm involved in hiring new programmers quite often. Way too many of them have absolutely no idea what they're doing, despite making it through some kind of program.
I guess maybe they made a lot of progress towards understanding the flavor of programming. Maybe they learned to leverage the "social dynamic" of programming to cobble together some garbage out of other people's code.
I think it would be better if they learned how to program, and had to prove they could do it before someone gave them a certification. You can learn programming the same way you learn anything else, and there's no reason to teach it or evaluate it differently. It's not magic, and I think with time and a sane approach you could teach most people how to do it in a couple years.
So because I can tell the difference between a neural net and "rudimentary genetic algorithms", I'm a Google shill. Sure. Great. Or I'm a troll. Sure. I'm an awful human being. OK. I am actually feeling sort of mean today.
But I'm bored. What the heck, I'll make an honest approach at a discussion.
I have spent a reasonable amount of time doing modelling with neural networks (and other ML technologies). I've had my largest successes with CNTK. Models built there became part of a property valuation project that has been very successful. These are all being maintained by other programmers now (in a company that was spin off to handle our "data provider" business), and have progressed beyond my level of expertise - but I retain some interest in how they're progressing and what new ideas and data sources are coming in.
I haven't used TensorFlow much, but I understand enough of it (and have enough background in the theory) that I can follow Google's paper (https://arxiv.org/pdf/1611.01578.pdf). I couldn't replicate their approach by any stretch - the models I've made have always had much simpler problem spaces and outputs, and I'm unclear on how a few things fit together here. But I have a general idea of what they're doing. To the point at hand - unless they are just lying about everything - then what they're doing clearly involves using a neural net to design other neural nets, and is not a "rudimentary genetic algorithm".
On that side of things, I learned about genetic algorithms in school, and have played with them enough to have a sense of how they work. They aren't my specialty by any stretch, but I do have some experience with them from a few years ago. As part of a competition, I made a program to play a simplified version of Poker against set-mixed-strategy opponents. The genetic algorithm I tried didn't end up working well; there wasn't enough time for the software to explore the space and find opponent's weaknesses (since while it was doing this, it wasn't winning) - it was stronger to just play well from the start, and my final submission in the competition was pretty lame (even though it performed well enough to get me to the next round). Kind of disappointing, but I felt like I learned something - and may have been able to do better with more programming time (or likely some big idea I missed).
Anyway, if you were saying that Support Vector Machines are like neural nets, I'd be like, "yeah, I see where you're coming from". There's similarity in the learning algorithms. They both employ similar methods of (and derive a lot of their power from) regularization. If you were saying that genetic algorithms and simulated annealing (something I have done a lot of) were similar, I'd say "yeah, sure, I get what you mean". They share the same kind of loosely directed iterative improvement. I could imagine reusing the same core in trying both those strategies.
But I don't see what you're getting at comparing neural nets and genetic algorithms (other than that, clearly, they could be employed against similar problems) nor do I see your point in dismissing Google's research here by saying it's the latter. My snap judgement from reading your first post was that you just hadn't read the article. When you persisted, I figured you were hazy on the subject matter, and were just sticking to your guns to save face.
But maybe I've misjudged and you have some insight here I'm missing.
Sorry, but do you really - deep down inside - think you're tricking me or something here? It's dumb. You got caught with your hand in the cookie jar here making up nonsense. Your bafflegab and sad, sad appeal to your own authority... I don't know what to say... it's just sad.
So who are you talking to? There's nobody else reading some weird side thread on an old Slashdot story.
The only other person you might be trying to convince is yourself. But, again, why? I don't care about the subject here. I don't care what you think about AI. I am not trying to win an argument because there is no argument. No rational person has any dispute about who's right here.
I just am anthropologically curious why you would continue to pretend like this? Is this how you conduct your life? Do you actually have a programming job you somehow bluster through? I mean, I quite often am interviewing programmers who somehow have held down jobs despite having no idea what they're doing. Isn't that kind of scary, going through life always worried you'll be exposed?
Why not just take a minute and actually learn what a neural net is and what a genetic algorithm is? Neither are really complicated concepts. Who knows, you might find it interesting.
Neural nets are not some fuzzy concept - it's a super clear set of data structures and algorithms, the basics of which have been in use for decades. The structure is super distinctive and easy to identify. It is not some accomplishment to make a neural net (it's something routinely done in introductory programming courses) nor is there any reason - any reason whatsoever - to question whether Google is using a neural net here. They build and use neural nets all the time, as do countless other companies and programmers in all sorts of situations. It is not notable to build or use a neural net, and there would be no reason to lie about it.
I have no idea what you think a neural net is, or why you doubt this is one. Your bizarre assertion that it's "some genetic algorithms" is nonsense. A genetic algorithm works in a very clear way - this is once again "intro to programming" level material - and that way is not similar to how a neural net works. Nobody who has any idea what these two things are would ever confuse them. You would never have a bunch of "genetic algorithms" that you decided to then call a neural net, because they are not similar concepts.
Pretty much everything you've said is bafflingly wrong nonsense. It is super clear you have no idea what you're talking about. Why not just stop?
Both the parent and child here are absolutely, obviously neural nets. Nobody should be contending this. They have nodes and weights and biases and output nodes - heck, you can see the network topologies (for the child NN) in the blog article.
Whatever you think about the utility of a neural net, or whatever you think about Google AI research, surely on reflection you have to agree that there's a neural net there? Right? That's like, clear unambiguous objective fact.
I think you've misunderstood what they're doing (possibly based on the terrible summary/article).
They are not testing all combinations of architecture in some kind of brute force - nor are they using an evolutionary algorithm (as many other posters here have assumed). Rather, a neural net is proposing new child architectures, and improving those architectures based on feedback (the performance of the child). Not only did it do well in design, but some of the features of its successful models were unexpected/novel. So it might be useful in and of itself (as network design is hard), and/or it might point towards new ideas in network design.
The summary and article are dumb hyperbole, but the actual blog article is pretty clear, their conclusion doesn't overstep their result, and in the end it seems like potentially important work they're doing.
Look, I agree it's dumb to call this "an AI" as the summary/pop-science article do. That implies something different than what's going on here. If you look at the research blog, they describe what they're making as "models" and "neural nets", and that's clearly a better description - and one that doesn't carry the same baggage as "making an AI" does.
However, I think it's very reasonable to describe what they're doing as artificial intelligence research (though clearly we remain a long way away from any sort of strong and/or general AI).
Leaving "AI" to the side, though, hopefully we can at least agree that what they're doing is not an evolutionary algorithm, as GP claimed.
You're wrong, and clearly didn't even read their summary - they specifically mention how this new approach (using a neural net to design neural nets) is performing better than previous attempts using evolutionary algorithms.
I take it you don't like Google, but they're doing probably the best work right now in the field of AI (and yes, this is AI research as defined by anyone other than pedants with axes to grind).
It's actually a lot more feasible now for a business to have all users Linux than it was in 2003 - with so many business apps moved to the web, desktop matters much less now than it used to. You could probably run a lot of businesses off a Chrome Book these days.
But yeah, the Linux desktop is still a straight-up terrible experience for a "regular user", if you're wondering. I try a new one every few years, and every time I try to evaluate them fresh - like, "last time I couldn't get Samba config to work with the GUI... but I'll try again now". This time, I hit a bug very early on (in Ubuntu?) where I tried to install a package (a.deb maybe?) by double clicking on it. It put a little icon up that said "waiting to install" that never went away and never did anything. I checked online, saw a bunch of bug reports dating back years.
One answer was something about "use the new Ubuntu Software Center to install the old Ubuntu Software, then open it with that", another said to use a command line to install some other package manager. The real answer, of course, was to use the command line. That worked fine for me, but the fact that that's an acceptable answer - one that nobody cares to fix - is kind of a summary of why desktop Linux is never going to get there.
When they kept pushing Kinect during the XBOne launch, I assumed they would eventually pair it with a VR headset. A VR system that knows where your legs, arms (and perhaps even fingers) are could have been really cool. Stomping things with your Godzilla legs or whatever could have been pretty neat.
Barring VR, they needed more creative experiences instead of using the Kinect as a terrible controller. Kinect Party was good. A simple "use your body to animate characters on screen" game would have been great too (I think there was one game where you could do this a bit in, like, a title screen, and that was the best part of that game). Just a 3d painting game (as we got later in Google Tilt Brush) would have been big.
There was a ton of Kinect games that my kids were straight-up not able to play - it felt like you had to hack through 3 layers of "anti-child-security" to even get some of them started (to be clear, these were games aimed at kids), then at some point it would fail to detect player and drop out.
But yeah, Kinect Party (after the rename, I think it was called something else previously) was the most fun my (at the time very young) kids ever had with a video game. It's bizarre that it wasn't duplicated, as it was the only thing that ever made the Kinect look good.
I've decided that this accomplishment -- a dizzying milestone in artificial intelligence that not long ago was though impossible or at least decades away -- is actually meaningless and doesn't prove anything and they should clearly have been working on some other problem. I have no idea how their system works, but I'm confident that their approach is just "brute force" (or something, I clearly have no idea what even that means) and won't generalize to any "real" problem solving (with my definition of "real problem" subject to change without notice).
I will only admit that any progress has been made towards artificial intelligence when computers perform exactly equivalent to humans in all tasks with no human intervention. I mean, I won't really, because I have weird quasi-spiritual hangups about believing computers can be intelligent, but that's where I'm putting the goal posts for now. Digital computers can't think, but I can because reasons. Free will or quantum mechanics or something else that I haven't thought about at all, probably.
Also, cotton gins and blacksmiths, therefore computers will never take our jobs. Amen.
To be clear, I would prefer to live in a world where more people felt free to discuss issues like this, and where unpopular opinions were met with much calmer reason and argument rather than blanket dismissals and insults. But that is not the current state of things; even if this guy's manifesto was much better written and well reasoned than it is, I think it would have met with effectively the same response. That's what I mean when I say this isn't currently a tenable position; just based on the subject matter and the extent to which it varies from the orthodoxy, he had no chance of successfully defending it.
I'm glad that society is advancing on many fronts in terms of equality, but I don't like how difficult/dangerous it is for reasonable people to disagree on and discuss this kind of issue with any kind of openness. I think it slows progress on all sorts of fronts, and drives many people in the "center" people towards worse decisions (like electing Trump). I don't know how to fix or change this state of affairs.
If he was smarter, or thought about it longer, he wouldn't have published it at all - it was clearly going to end with him being either ignored or fired. I don't think it would matter if it was a little better baked, it's just not a tenable basic position right now.
And that's the problem I see here - the only people willing to discuss these kinds of issues are: 1. Those whose opinions lean a certain way, conforming to a narrow current orthodoxy 2. Those in "the middle" who aren't smart enough to see how badly this is going to end for themselves (eg. this guy) 3. Actual bigots, who don't mind being called out as such because they think bigotry is right
I actually don't have much problem with "the orthodoxy" here in terms of position - my own views are left leaning enough that I could safely express them in most forums. But I am bothered at the current impulse to just kind of "make unspeakable" a bunch of possible other positions. It doesn't end well. It drives people to extremes. I think it's how Trump got elected. Media just kept telling a class of people (representing a large percentage of Americans): your concerns aren't real, you're a degenerate if you don't immediately agree with this whole list of things. It's no surprise they listened to a guy who acknowledged them and the things they're worried about (whether your or I might think those things are valid or not).
Now, again, clearly this guy at Google didn't think things out well. But it's also fairly clear he's not alone in his opinions, and it's not like some smarter representative is going to come forward to better articulate that group's position and start some productive dialog. Why would someone thoughtful stick their neck out like that? I don't see it happening. But without some kind of representation for this group - without allowing people to express opinions somewhere in this middle - we're going to lose a lot of these people to actual bigotry, and a lot of problems will go unsolved.
Trump has not been doing well on the current political fronts (hilariously poor speeches, Russia collusion), so it's a good call for him to shift discussion. To be clear, I think the Russia stuff is mostly a matter of stupidity/greed (with a healthy dose of anachronistic RED SCARE) rather than high treason or something, but nonetheless these stories are not going in Trump's favor (and his previous attempts at handling/distraction have been super feeble, eg. Clinton has Russia ties too!)..so it's a good idea for him to shift the discussion to a more traditional right-left battlefield: a moral panic. Here, his part is much simpler to play: "oh no, the ~new thing~ is going to destroy the foundation of our society/military/families" - and this particular subject is great because the left will respond predictably in a way that will alienate around half the country. From Trump's perspective, much better to focus on a polarizing issue where he has 45% support and a clear game plan, than let focus stay on his bumbling kids and their stupid Russian entanglements.
Zuckerberg is actually a pretty good programmer. You can still find some of his old submitted code on TopCoder. He wasn't, like, a super serious competitor - and you can't credit Facebook's success to his unworldly programming skills or something - but he has some very reasonable tech background and skills.
Yeah, see I think fruit fails as "medicine" not because it's preventative, but rather because I wouldn't call it a "drug or preparation". If you, say, ground up tree bark and snorted it to prevent a cold, I would have no problem calling that a medicine. I mean, there's foggy areas if you want there to be (oh, this "baked potato" is medicine because it's a "baked preparation" that I use to prevent the disease of "starvation"), but I think that definition is actually pretty solid in terms of matching how people use the word.
And lots of perfectly reasonable medicine is preventative in just the way you seem to be against. I have no qualms with, say, getting vaccinated - even for a disease that I'm very unlikely to encounter. Different preventative medicines will have different pros and cons. And while vaccines might not spring to mind if I was asked to give examples of "medication", I do think it's reasonable to call them that.
I agree that calling antibacterial soap an "antibiotic" doesn't fit how that word is used - but trying to suggest that's because it isn't "medicine" doesn't make sense.
In your own linked dictionary, the relevant definition for medicine is: "A drug or other preparation for the treatment or prevention of disease." That definition would obviously include antibacterial soaps which are preparations used to prevent a disease. I can imagine being prescribed an antibacterial soap and calling it medicine (but, again, on the flipside, I agree that doctors would not call such a treatment an antibiotic - though I can't point to a specific reason they wouldn't).
VR isn't perfectly standardized, but SteamVR (and most VR games on Steam) support both the Rift and the Vive, and do so in a way that's mostly transparent as a player. Overall, there isn't really that much difference at this point in features (and many people prefer the Touch controllers to the Vive wands) - so at this price, I think the Oculus is the clear choice (I say this despite buying a Vive on launch and having got a lot of fun out of it).
Slashdot Mirror
Remember? The one where computers performed at par with humans on a very closely related task to this?
Like, I understand why you're aroused by this impossible-to-overcome flaw you're imagining, where computers could never guess what the "it" is referring to - but it turns out they don't work like the text parsers in 80s chat-bots, and are capable of correctly interpreting these references. They continue to improve at exactly this kind of work.
And hey, do you know what the "deep" in "deep neural net" means? It means there's more layers of nodes between input and output. It's not something metaphysical or subjective - if you look at the diagram of how nodes are connected you can describe how deep it is. Do you object when someone calls a swimming pool "deep"? That's pretty much the same mistake you're making.
And is this thing like a brain? Well, it can do some of the same tasks, and some of them better. Does it use the same mechanisms to do that work? The answer is somewhere between "obviously not" and "who cares?". Planes don't fly the same way birds do - but they were inspired by birds and we call the jutty-out parts "wings". Neural nets were inspired by our (limited) understanding of neurons - and it turns out that (just like wings) they don't have to work just the same way in order to create useful results. How would you respond to someone on an aviation board saying "Those things don't work anything like real wings, they're nothing like birds"? Do you see how, while true, that's kind of not an interesting point to make on every article?
In the end, the key question about a plane is whether it flies - whether it does the thing it was made to do. And neural nets are getting better at flying all the time.
I think you're right that "car ownership" would become less accessible over time, but I think "transportation by car" is likely to get very cheap. I think we'll see ubiquitous "Driverless Uber" style services pop up everywhere (in a variety of flavors), and that's how the non-wealthy get around (at least in urban settings).
But I also think the transition will take a while, especially in rural areas - enough so that I don't think the transition will be too brutal for most people economically. If anything, some may benefit from cheaper human driven cars as they're essentially discarded. But yeah, further out in time it will definitely not be good for people who like driving - just as it's currently very expensive/awkward to maintain and use a horse drawn carriage.
I think some people are jumping the gun a bit right now, but once driverless cars are reliable they're going to go from niche to everywhere very quickly.
I never open the Youtube app, and only end up on Youtube through a link.
And I never want to open random links to Youtube (or Twitter or Facebook - or any of the preloaded apps I can't remove) with some dumb special app - the browser version is fine, doesn't force me to log in, and stays in its tab where I put it.
Hopefully Google gets in some fight with everyone else (and, uh, itself) so that all this garbage stays in browser tabs where it belongs.
I think a lot of people are curious about blockchain technology because Bitcoin has been in the news - people don't want to miss out on the new thing in technology. It's natural that IBM and Oracle are going to use a wedge like that to try to get in the door with people, but I don't think that many businesses are going to have problems where blockchain tech is an important part of the solution.
I'm involved in hiring new programmers quite often. Way too many of them have absolutely no idea what they're doing, despite making it through some kind of program.
I guess maybe they made a lot of progress towards understanding the flavor of programming. Maybe they learned to leverage the "social dynamic" of programming to cobble together some garbage out of other people's code.
I think it would be better if they learned how to program, and had to prove they could do it before someone gave them a certification. You can learn programming the same way you learn anything else, and there's no reason to teach it or evaluate it differently. It's not magic, and I think with time and a sane approach you could teach most people how to do it in a couple years.
So because I can tell the difference between a neural net and "rudimentary genetic algorithms", I'm a Google shill. Sure. Great. Or I'm a troll. Sure. I'm an awful human being. OK. I am actually feeling sort of mean today.
But I'm bored. What the heck, I'll make an honest approach at a discussion.
I have spent a reasonable amount of time doing modelling with neural networks (and other ML technologies). I've had my largest successes with CNTK. Models built there became part of a property valuation project that has been very successful. These are all being maintained by other programmers now (in a company that was spin off to handle our "data provider" business), and have progressed beyond my level of expertise - but I retain some interest in how they're progressing and what new ideas and data sources are coming in.
I haven't used TensorFlow much, but I understand enough of it (and have enough background in the theory) that I can follow Google's paper (https://arxiv.org/pdf/1611.01578.pdf). I couldn't replicate their approach by any stretch - the models I've made have always had much simpler problem spaces and outputs, and I'm unclear on how a few things fit together here. But I have a general idea of what they're doing. To the point at hand - unless they are just lying about everything - then what they're doing clearly involves using a neural net to design other neural nets, and is not a "rudimentary genetic algorithm".
On that side of things, I learned about genetic algorithms in school, and have played with them enough to have a sense of how they work. They aren't my specialty by any stretch, but I do have some experience with them from a few years ago. As part of a competition, I made a program to play a simplified version of Poker against set-mixed-strategy opponents. The genetic algorithm I tried didn't end up working well; there wasn't enough time for the software to explore the space and find opponent's weaknesses (since while it was doing this, it wasn't winning) - it was stronger to just play well from the start, and my final submission in the competition was pretty lame (even though it performed well enough to get me to the next round). Kind of disappointing, but I felt like I learned something - and may have been able to do better with more programming time (or likely some big idea I missed).
Anyway, if you were saying that Support Vector Machines are like neural nets, I'd be like, "yeah, I see where you're coming from". There's similarity in the learning algorithms. They both employ similar methods of (and derive a lot of their power from) regularization. If you were saying that genetic algorithms and simulated annealing (something I have done a lot of) were similar, I'd say "yeah, sure, I get what you mean". They share the same kind of loosely directed iterative improvement. I could imagine reusing the same core in trying both those strategies.
But I don't see what you're getting at comparing neural nets and genetic algorithms (other than that, clearly, they could be employed against similar problems) nor do I see your point in dismissing Google's research here by saying it's the latter. My snap judgement from reading your first post was that you just hadn't read the article. When you persisted, I figured you were hazy on the subject matter, and were just sticking to your guns to save face.
But maybe I've misjudged and you have some insight here I'm missing.
Sorry, but do you really - deep down inside - think you're tricking me or something here? It's dumb. You got caught with your hand in the cookie jar here making up nonsense. Your bafflegab and sad, sad appeal to your own authority... I don't know what to say... it's just sad.
So who are you talking to? There's nobody else reading some weird side thread on an old Slashdot story.
The only other person you might be trying to convince is yourself. But, again, why? I don't care about the subject here. I don't care what you think about AI. I am not trying to win an argument because there is no argument. No rational person has any dispute about who's right here.
I just am anthropologically curious why you would continue to pretend like this? Is this how you conduct your life? Do you actually have a programming job you somehow bluster through? I mean, I quite often am interviewing programmers who somehow have held down jobs despite having no idea what they're doing. Isn't that kind of scary, going through life always worried you'll be exposed?
Why not just take a minute and actually learn what a neural net is and what a genetic algorithm is? Neither are really complicated concepts. Who knows, you might find it interesting.
Neural nets are not some fuzzy concept - it's a super clear set of data structures and algorithms, the basics of which have been in use for decades. The structure is super distinctive and easy to identify. It is not some accomplishment to make a neural net (it's something routinely done in introductory programming courses) nor is there any reason - any reason whatsoever - to question whether Google is using a neural net here. They build and use neural nets all the time, as do countless other companies and programmers in all sorts of situations. It is not notable to build or use a neural net, and there would be no reason to lie about it.
I have no idea what you think a neural net is, or why you doubt this is one. Your bizarre assertion that it's "some genetic algorithms" is nonsense. A genetic algorithm works in a very clear way - this is once again "intro to programming" level material - and that way is not similar to how a neural net works. Nobody who has any idea what these two things are would ever confuse them. You would never have a bunch of "genetic algorithms" that you decided to then call a neural net, because they are not similar concepts.
Pretty much everything you've said is bafflingly wrong nonsense. It is super clear you have no idea what you're talking about. Why not just stop?
Both the parent and child here are absolutely, obviously neural nets. Nobody should be contending this. They have nodes and weights and biases and output nodes - heck, you can see the network topologies (for the child NN) in the blog article.
Whatever you think about the utility of a neural net, or whatever you think about Google AI research, surely on reflection you have to agree that there's a neural net there? Right? That's like, clear unambiguous objective fact.
I think you've misunderstood what they're doing (possibly based on the terrible summary/article).
They are not testing all combinations of architecture in some kind of brute force - nor are they using an evolutionary algorithm (as many other posters here have assumed). Rather, a neural net is proposing new child architectures, and improving those architectures based on feedback (the performance of the child). Not only did it do well in design, but some of the features of its successful models were unexpected/novel. So it might be useful in and of itself (as network design is hard), and/or it might point towards new ideas in network design.
The summary and article are dumb hyperbole, but the actual blog article is pretty clear, their conclusion doesn't overstep their result, and in the end it seems like potentially important work they're doing.
Look, I agree it's dumb to call this "an AI" as the summary/pop-science article do. That implies something different than what's going on here. If you look at the research blog, they describe what they're making as "models" and "neural nets", and that's clearly a better description - and one that doesn't carry the same baggage as "making an AI" does.
However, I think it's very reasonable to describe what they're doing as artificial intelligence research (though clearly we remain a long way away from any sort of strong and/or general AI).
Leaving "AI" to the side, though, hopefully we can at least agree that what they're doing is not an evolutionary algorithm, as GP claimed.
You're wrong, and clearly didn't even read their summary - they specifically mention how this new approach (using a neural net to design neural nets) is performing better than previous attempts using evolutionary algorithms.
I take it you don't like Google, but they're doing probably the best work right now in the field of AI (and yes, this is AI research as defined by anyone other than pedants with axes to grind).
It's actually a lot more feasible now for a business to have all users Linux than it was in 2003 - with so many business apps moved to the web, desktop matters much less now than it used to. You could probably run a lot of businesses off a Chrome Book these days.
But yeah, the Linux desktop is still a straight-up terrible experience for a "regular user", if you're wondering. I try a new one every few years, and every time I try to evaluate them fresh - like, "last time I couldn't get Samba config to work with the GUI... but I'll try again now". This time, I hit a bug very early on (in Ubuntu?) where I tried to install a package (a .deb maybe?) by double clicking on it. It put a little icon up that said "waiting to install" that never went away and never did anything. I checked online, saw a bunch of bug reports dating back years.
One answer was something about "use the new Ubuntu Software Center to install the old Ubuntu Software, then open it with that", another said to use a command line to install some other package manager. The real answer, of course, was to use the command line. That worked fine for me, but the fact that that's an acceptable answer - one that nobody cares to fix - is kind of a summary of why desktop Linux is never going to get there.
When they kept pushing Kinect during the XBOne launch, I assumed they would eventually pair it with a VR headset. A VR system that knows where your legs, arms (and perhaps even fingers) are could have been really cool. Stomping things with your Godzilla legs or whatever could have been pretty neat.
Barring VR, they needed more creative experiences instead of using the Kinect as a terrible controller. Kinect Party was good. A simple "use your body to animate characters on screen" game would have been great too (I think there was one game where you could do this a bit in, like, a title screen, and that was the best part of that game). Just a 3d painting game (as we got later in Google Tilt Brush) would have been big.
There was a ton of Kinect games that my kids were straight-up not able to play - it felt like you had to hack through 3 layers of "anti-child-security" to even get some of them started (to be clear, these were games aimed at kids), then at some point it would fail to detect player and drop out.
But yeah, Kinect Party (after the rename, I think it was called something else previously) was the most fun my (at the time very young) kids ever had with a video game. It's bizarre that it wasn't duplicated, as it was the only thing that ever made the Kinect look good.
Uh... dude... I was making fun of you. Well, you and Gweihir (sic?), who I assume is taking a day off to scream at pigeons.
I've decided that this accomplishment -- a dizzying milestone in artificial intelligence that not long ago was though impossible or at least decades away -- is actually meaningless and doesn't prove anything and they should clearly have been working on some other problem. I have no idea how their system works, but I'm confident that their approach is just "brute force" (or something, I clearly have no idea what even that means) and won't generalize to any "real" problem solving (with my definition of "real problem" subject to change without notice).
I will only admit that any progress has been made towards artificial intelligence when computers perform exactly equivalent to humans in all tasks with no human intervention. I mean, I won't really, because I have weird quasi-spiritual hangups about believing computers can be intelligent, but that's where I'm putting the goal posts for now. Digital computers can't think, but I can because reasons. Free will or quantum mechanics or something else that I haven't thought about at all, probably.
Also, cotton gins and blacksmiths, therefore computers will never take our jobs. Amen.
To be clear, I would prefer to live in a world where more people felt free to discuss issues like this, and where unpopular opinions were met with much calmer reason and argument rather than blanket dismissals and insults. But that is not the current state of things; even if this guy's manifesto was much better written and well reasoned than it is, I think it would have met with effectively the same response. That's what I mean when I say this isn't currently a tenable position; just based on the subject matter and the extent to which it varies from the orthodoxy, he had no chance of successfully defending it.
I'm glad that society is advancing on many fronts in terms of equality, but I don't like how difficult/dangerous it is for reasonable people to disagree on and discuss this kind of issue with any kind of openness. I think it slows progress on all sorts of fronts, and drives many people in the "center" people towards worse decisions (like electing Trump). I don't know how to fix or change this state of affairs.
If he was smarter, or thought about it longer, he wouldn't have published it at all - it was clearly going to end with him being either ignored or fired. I don't think it would matter if it was a little better baked, it's just not a tenable basic position right now.
And that's the problem I see here - the only people willing to discuss these kinds of issues are:
1. Those whose opinions lean a certain way, conforming to a narrow current orthodoxy
2. Those in "the middle" who aren't smart enough to see how badly this is going to end for themselves (eg. this guy)
3. Actual bigots, who don't mind being called out as such because they think bigotry is right
I actually don't have much problem with "the orthodoxy" here in terms of position - my own views are left leaning enough that I could safely express them in most forums. But I am bothered at the current impulse to just kind of "make unspeakable" a bunch of possible other positions. It doesn't end well. It drives people to extremes. I think it's how Trump got elected. Media just kept telling a class of people (representing a large percentage of Americans): your concerns aren't real, you're a degenerate if you don't immediately agree with this whole list of things. It's no surprise they listened to a guy who acknowledged them and the things they're worried about (whether your or I might think those things are valid or not).
Now, again, clearly this guy at Google didn't think things out well. But it's also fairly clear he's not alone in his opinions, and it's not like some smarter representative is going to come forward to better articulate that group's position and start some productive dialog. Why would someone thoughtful stick their neck out like that? I don't see it happening. But without some kind of representation for this group - without allowing people to express opinions somewhere in this middle - we're going to lose a lot of these people to actual bigotry, and a lot of problems will go unsolved.
Just got an e-mail, was modded "Troll". At least some parts of Slashdot never change.
Trump has not been doing well on the current political fronts (hilariously poor speeches, Russia collusion), so it's a good call for him to shift discussion. To be clear, I think the Russia stuff is mostly a matter of stupidity/greed (with a healthy dose of anachronistic RED SCARE) rather than high treason or something, but nonetheless these stories are not going in Trump's favor (and his previous attempts at handling/distraction have been super feeble, eg. Clinton has Russia ties too!) ..so it's a good idea for him to shift the discussion to a more traditional right-left battlefield: a moral panic. Here, his part is much simpler to play: "oh no, the ~new thing~ is going to destroy the foundation of our society/military/families" - and this particular subject is great because the left will respond predictably in a way that will alienate around half the country. From Trump's perspective, much better to focus on a polarizing issue where he has 45% support and a clear game plan, than let focus stay on his bumbling kids and their stupid Russian entanglements.
Zuckerberg is actually a pretty good programmer. You can still find some of his old submitted code on TopCoder. He wasn't, like, a super serious competitor - and you can't credit Facebook's success to his unworldly programming skills or something - but he has some very reasonable tech background and skills.
Yeah, see I think fruit fails as "medicine" not because it's preventative, but rather because I wouldn't call it a "drug or preparation". If you, say, ground up tree bark and snorted it to prevent a cold, I would have no problem calling that a medicine. I mean, there's foggy areas if you want there to be (oh, this "baked potato" is medicine because it's a "baked preparation" that I use to prevent the disease of "starvation"), but I think that definition is actually pretty solid in terms of matching how people use the word.
And lots of perfectly reasonable medicine is preventative in just the way you seem to be against. I have no qualms with, say, getting vaccinated - even for a disease that I'm very unlikely to encounter. Different preventative medicines will have different pros and cons. And while vaccines might not spring to mind if I was asked to give examples of "medication", I do think it's reasonable to call them that.
I agree that calling antibacterial soap an "antibiotic" doesn't fit how that word is used - but trying to suggest that's because it isn't "medicine" doesn't make sense.
In your own linked dictionary, the relevant definition for medicine is: "A drug or other preparation for the treatment or prevention of disease." That definition would obviously include antibacterial soaps which are preparations used to prevent a disease. I can imagine being prescribed an antibacterial soap and calling it medicine (but, again, on the flipside, I agree that doctors would not call such a treatment an antibiotic - though I can't point to a specific reason they wouldn't).
VR isn't perfectly standardized, but SteamVR (and most VR games on Steam) support both the Rift and the Vive, and do so in a way that's mostly transparent as a player. Overall, there isn't really that much difference at this point in features (and many people prefer the Touch controllers to the Vive wands) - so at this price, I think the Oculus is the clear choice (I say this despite buying a Vive on launch and having got a lot of fun out of it).