Japanese as an interims [sic, interlingua] language? I don't think so, at least I've never seen a system that did that. Citation? (See e.g. this 1998 paper, of which Alex Waibel is a co-author: isl.anthropomatik.kit.edu/cmu-kit/english/5633.php; no mention of Japanese as an interlingua; and the Kauers, Vogel, Fügen, and Waibel paper in INTERSPEECH, 2002 doesn't mention Japanese at all.) Claims that language X is better for clearly/ unambiguously depicting thought are usually made by native speakers of that language, and are probably as valid as claims that language Y (Hebrew, Tamil, Arabic...) is the ancestral language of all others.
I have not read the original article, so take my comments with some grains of salt.
But speaking as one who once wrote a syntactic grammar for a parser of English (still in use by a large manufacturer 30 years later, albeit in modified form), the problem with rule-based grammars that lack any statistical weights is that they come up with an unbelievably large number of parses for many real-world sentences. The problem is then to find which of those parses is the correct one, and that's what statistical weights (or some other kind of heuristic) allow you to do. Naturally the weights aren't perfect; as I see it, they're substituting for a real semantic and pragmatic analysis that would allow you to rule out nonsensical parses (e.g. the parse of "time flies like an arrow" that would give a semantic representation corresponding to the normal sense of "fruit flies like a banana"). The weights are lexically driven, which means they tend to be more domain-specific than the actual rules of a rule-based grammar.
It's therefore hard to compare a rule-based parser with a statistical parser learned from a treebank, unless you specify whether the rule-based parser is being graded for whether the "correct" parse appears *among* its many parses, or it's being graded for having the correct parse as its one-best parse. There's also a ceiling for measured performance, which has to do with the inter-annotator agreement in the treebank that you're using for testing (and by the fact that some sentences are ambiguous even in context).
Apropos of this, the VISL article has a caveat at the end: "...our evaluation was less rigid than optimally desirable, since it did not use multiple annotators and manual revision was performed on top of an automatic analysis, potentially creating a parser-friendly bias in ambiguous cases." So the 96% accuracy claim is suspect, not to mention that a comparison of the Google system is already difficult because Spanish =/= English. (Spanish has more morphology on verbs, it's pro-drop, it has relatively free word order compared to English,...)
So I don't believe you can say that "Google is hopelessly behind the state of the art."
The latest victim (besides you and me, that is) is Adobe Acrobat--DC, I think is the version where it went to pot. Now there are two rows of information at the top that in any normal app would have one row, in a reasonably sized black font. But Adobe has to be different, so both rows are in a huge gray (= low contrast) font (or huge gray icons), with a huge amount of wasted white space. If you don't believe me (ok, AC, I think you would, but if other readers don't believe me), then do a web search for "Adobe Acrobat DC ugly".
Yeah, I noticed this awhile back. And I got ticked off at Google when they made their last change to Maps a year or two ago, after they had royally messed up News a few years before that (and ignored unanimous user complaints on their feedback pages). So I tried Bing and Yahoo. Unfortunately, for the things I search for (linguistics and computational linguistics articles, programming hacks), neither Bing nor Yahoo was anywhere near as good as Google. So reluctantly, I went back to Google. https://www.youtube.com/watch?...
"...holding a book at full arms length and then complaining about the size of the text": Umm, I started holding books at arms length when I was in my forties. Now I have to hang them on the wall on the other side of the room. Ok, I can use those new-fangled pieces of glass that you stick in front of your eyes, and get a little closer.
It seems to still be running on my computer. And maybe I'll switch to Chrome when they put a real menu on it. (In Windows. I understand that in the MacOS, it does have a real menu.)
If I were to define the best SciFi book, it would be the Analog magazine, at least when John Campbell was in charge. And since a lot of SciFi fans will doubtless be reading this, I'm going to take the opportunity to point out that Project Gutenberg (a project which has been active since the early days of the internet--in fact, the ArpaNet) has digitized a bunch of articles from Analog magazine, some dating back to the 1930s (but especially important for people like me, many from the 1960s). You can find these, along with articles from several other SciFi magazines, at http://www.freesfonline.de/Mag....
I believe this theme has been explored in SciFi (Elysium?), and while the Fi in SciFi stands for "Fiction", I think it brings up legitimate concerns.
A post-manufacturing (at least by humans) world is the backdrop of Stanislaw Lem's "Return from the Stars" = Powrót z gwiazd (ok, I copy-pasted that from the wikipedia). The people seem very busy, but it's not clear what they do.
I think I know why it's so primitive looking. When I was a kid, there was Mad Magazine. And I remember a faux letter to the editor that started out s.t. like "I'm writing this letter in crayon, because they won't let me have sharp instruments here..." I think that writer must now be working for Microsoft.
I haven't retired yet (maybe by the age of 80 I'll be able to), and like you I use TBird at home, but I (still) have to use Outlook at work. And like you, for most purposes I prefer TBird. I guess the biggest reason is a plugin called Nostalgy, that makes filing emails *so* much easier. Nothing like it for Outlook, afaik, so I spend half my day (it seems) scrolling up and down through my Outlook folders. But I will say that search in Outlook is much better than in TBird. Search in TBird is excruciatingly slow, so slow that I downloaded the (free) MailStore app, which makes searches almost instantaneous. If there were a way to integrate (as in "put both into a single program") MailStore and TBird, it would leave Outlook in the dust.
Btw, don't get me started about that *&*--!^#@ ribbon... The only thing worse is the new interface in Adobe Acrobat DC. (So I use PDF-SChange Editor at home.)
Strange, I have one sitting right in front of me, and it seems to still be alive. A few hours ago, I was sitting next to my son at dinner, and he was showing me his Android. I used to have an Android, but I much prefer the Windows OS. He said, "But you can't get App A on Windows!" I opened up the Windows phone store, and downloaded App A. "Oh, but you can't get App B!" I went back to the Windows store, got App B.
Obviously there are far more apps for Android than there are for Windows. Some would say I was just lucky. I'd say I don't need the junky apps that mostly fill the Android store. In all the years I had an Android phone, I never found as good a weather app as the one that came built into my Windows phone. I prefer the navigation apps I have on Windows to the ones I had on my Android phone (although I hear that app maker has jumped ship). It's much easier to set the alarm on my Windows phone (the Android phone was always over-shooting). And in general I find my Windows phone easier to use.
I am _not_ an Ms shill, and I never moved from Win7 to Win8 on my desktop (nor have I found any compelling reason yet to move to Win10). And for programming, Linux is far superior. But Android is, IMNSHO, a piece of junk.
I realize that barring some miracle, the Windows phone will probably be dead some day. But it isn't yet.
I have yet to figure out what this IoT is good for (I think I'm agreeing with you). So my toaster can text me to tell me my bagel is done? That's not high on my priority list... I suppose I could get a odor-of-rotting-food detector for my refr that would text me, but I can smell that just fine when I open the door. And I recently replaced the smart thermostats in my house with dumb ones, because the smart ones (not networked) were too hard to program. (I program computers as part of my job, so no, I'm not a technophobe.)
BTW, re toasters: my last one was B&D, and it was worse than any toaster I've ever had. Why, you ask? I'm so glad you asked! It was because both the automatic pop-up and the push-up lever pushed the toast up only about an inch. Barely sufficient for a full-sized piece of bread, and completely insufficient for a bagel or English muffin; you had to reach down in the toaster with a fork (or turn it upside down) to get those out. And no, that is _not_ safe.
"you have suddenly introduced a whole host of new variables" How about a controlled experiment? Send two identical machines up in a single launch; turn one of them on but not the other. If they begin separating in the expected direction, you have something. You could even switch the direction the running machine is pointed in, switch which one is turned on, etc.
I of course agree that we would want to find out how it works, if for no other reason than the fact that we might have built a very inefficient machine. But I think there's more than one way to do the science, and a controlled experiment out in space seems one perfectly reasonable way to do it.
Right, a huge diff--because in order to accelerate using propellant, at each instant you have to accelerate the propellant that you haven't used yet as well as the engines, power generating device, and the cargo. (You could use the propellant all at once, to some approximation, but the resulting acceleration would pretty much squish any cargo, human or otherwise.) And the propellant is a huge mass in comparison to the rest. (Mass of fully fueled Saturn V, 3 million kilos; mass of payload to lunar orbit: 50 thousand kilos.) With a propellant-less drive, you only need to accelerate the engines, power generating device, and cargo.
Now if the theory of how this drive works is correct, is there any way to shield the ship--or at least part of it (presumably not the engine) from the Unruh radiation? Thereby reducing its mass...
The headline is misleading; from the article: "There are currently no plans in the pipeline to clone and produce humans in a bid to eradicate disease, but Xiaochun has said that this can change if people become more open to the idea of it." Time for a Sixth Day law.
IMHO, this story (I'll assume it's true, and not just boasting or wishful thinking) deserves to be told in a much wider forum (like the Wall Street Journal), with the X, Y and Z replaced by real names, and especially with your company's name.
Slashdot Mirror
Japanese as an interims [sic, interlingua] language? I don't think so, at least I've never seen a system that did that. Citation? (See e.g. this 1998 paper, of which Alex Waibel is a co-author: isl.anthropomatik.kit.edu/cmu-kit/english/5633.php; no mention of Japanese as an interlingua; and the Kauers, Vogel, Fügen, and Waibel paper in INTERSPEECH, 2002 doesn't mention Japanese at all.) Claims that language X is better for clearly/ unambiguously depicting thought are usually made by native speakers of that language, and are probably as valid as claims that language Y (Hebrew, Tamil, Arabic...) is the ancestral language of all others.
Hear them down in Soho Square, dropping aitches everywhere, speaking English any way they please.
I have not read the original article, so take my comments with some grains of salt.
But speaking as one who once wrote a syntactic grammar for a parser of English (still in use by a large manufacturer 30 years later, albeit in modified form), the problem with rule-based grammars that lack any statistical weights is that they come up with an unbelievably large number of parses for many real-world sentences. The problem is then to find which of those parses is the correct one, and that's what statistical weights (or some other kind of heuristic) allow you to do. Naturally the weights aren't perfect; as I see it, they're substituting for a real semantic and pragmatic analysis that would allow you to rule out nonsensical parses (e.g. the parse of "time flies like an arrow" that would give a semantic representation corresponding to the normal sense of "fruit flies like a banana"). The weights are lexically driven, which means they tend to be more domain-specific than the actual rules of a rule-based grammar.
It's therefore hard to compare a rule-based parser with a statistical parser learned from a treebank, unless you specify whether the rule-based parser is being graded for whether the "correct" parse appears *among* its many parses, or it's being graded for having the correct parse as its one-best parse. There's also a ceiling for measured performance, which has to do with the inter-annotator agreement in the treebank that you're using for testing (and by the fact that some sentences are ambiguous even in context).
Apropos of this, the VISL article has a caveat at the end: "...our evaluation was less rigid than optimally desirable, since it did not use multiple annotators and manual revision was performed on top of an automatic analysis, potentially creating a parser-friendly bias in ambiguous cases." So the 96% accuracy claim is suspect, not to mention that a comparison of the Google system is already difficult because Spanish =/= English. (Spanish has more morphology on verbs, it's pro-drop, it has relatively free word order compared to English,...)
So I don't believe you can say that "Google is hopelessly behind the state of the art."
...or from aliens. No, not *that* kind of alien!
or onto those creeks that someone mentioned as being more visible than the roads.
The latest victim (besides you and me, that is) is Adobe Acrobat--DC, I think is the version where it went to pot. Now there are two rows of information at the top that in any normal app would have one row, in a reasonably sized black font. But Adobe has to be different, so both rows are in a huge gray (= low contrast) font (or huge gray icons), with a huge amount of wasted white space. If you don't believe me (ok, AC, I think you would, but if other readers don't believe me), then do a web search for "Adobe Acrobat DC ugly".
Yeah, I noticed this awhile back. And I got ticked off at Google when they made their last change to Maps a year or two ago, after they had royally messed up News a few years before that (and ignored unanimous user complaints on their feedback pages). So I tried Bing and Yahoo. Unfortunately, for the things I search for (linguistics and computational linguistics articles, programming hacks), neither Bing nor Yahoo was anywhere near as good as Google. So reluctantly, I went back to Google. https://www.youtube.com/watch?...
"...holding a book at full arms length and then complaining about the size of the text": Umm, I started holding books at arms length when I was in my forties. Now I have to hang them on the wall on the other side of the room. Ok, I can use those new-fangled pieces of glass that you stick in front of your eyes, and get a little closer.
Wait until you get old...
It seems to still be running on my computer. And maybe I'll switch to Chrome when they put a real menu on it. (In Windows. I understand that in the MacOS, it does have a real menu.)
Actually, Vivaldi looks pretty good.
If I were to define the best SciFi book, it would be the Analog magazine, at least when John Campbell was in charge. And since a lot of SciFi fans will doubtless be reading this, I'm going to take the opportunity to point out that Project Gutenberg (a project which has been active since the early days of the internet--in fact, the ArpaNet) has digitized a bunch of articles from Analog magazine, some dating back to the 1930s (but especially important for people like me, many from the 1960s). You can find these, along with articles from several other SciFi magazines, at http://www.freesfonline.de/Mag....
I believe this theme has been explored in SciFi (Elysium?), and while the Fi in SciFi stands for "Fiction", I think it brings up legitimate concerns.
A post-manufacturing (at least by humans) world is the backdrop of Stanislaw Lem's "Return from the Stars" = Powrót z gwiazd (ok, I copy-pasted that from the wikipedia). The people seem very busy, but it's not clear what they do.
I think I know why it's so primitive looking. When I was a kid, there was Mad Magazine. And I remember a faux letter to the editor that started out s.t. like "I'm writing this letter in crayon, because they won't let me have sharp instruments here..." I think that writer must now be working for Microsoft.
I haven't retired yet (maybe by the age of 80 I'll be able to), and like you I use TBird at home, but I (still) have to use Outlook at work. And like you, for most purposes I prefer TBird. I guess the biggest reason is a plugin called Nostalgy, that makes filing emails *so* much easier. Nothing like it for Outlook, afaik, so I spend half my day (it seems) scrolling up and down through my Outlook folders. But I will say that search in Outlook is much better than in TBird. Search in TBird is excruciatingly slow, so slow that I downloaded the (free) MailStore app, which makes searches almost instantaneous. If there were a way to integrate (as in "put both into a single program") MailStore and TBird, it would leave Outlook in the dust.
Btw, don't get me started about that *&*--!^#@ ribbon... The only thing worse is the new interface in Adobe Acrobat DC. (So I use PDF-SChange Editor at home.)
Strange, I have one sitting right in front of me, and it seems to still be alive. A few hours ago, I was sitting next to my son at dinner, and he was showing me his Android. I used to have an Android, but I much prefer the Windows OS. He said, "But you can't get App A on Windows!" I opened up the Windows phone store, and downloaded App A. "Oh, but you can't get App B!" I went back to the Windows store, got App B.
Obviously there are far more apps for Android than there are for Windows. Some would say I was just lucky. I'd say I don't need the junky apps that mostly fill the Android store. In all the years I had an Android phone, I never found as good a weather app as the one that came built into my Windows phone. I prefer the navigation apps I have on Windows to the ones I had on my Android phone (although I hear that app maker has jumped ship). It's much easier to set the alarm on my Windows phone (the Android phone was always over-shooting). And in general I find my Windows phone easier to use.
I am _not_ an Ms shill, and I never moved from Win7 to Win8 on my desktop (nor have I found any compelling reason yet to move to Win10). And for programming, Linux is far superior. But Android is, IMNSHO, a piece of junk.
I realize that barring some miracle, the Windows phone will probably be dead some day. But it isn't yet.
I have yet to figure out what this IoT is good for (I think I'm agreeing with you). So my toaster can text me to tell me my bagel is done? That's not high on my priority list... I suppose I could get a odor-of-rotting-food detector for my refr that would text me, but I can smell that just fine when I open the door. And I recently replaced the smart thermostats in my house with dumb ones, because the smart ones (not networked) were too hard to program. (I program computers as part of my job, so no, I'm not a technophobe.)
BTW, re toasters: my last one was B&D, and it was worse than any toaster I've ever had. Why, you ask? I'm so glad you asked! It was because both the automatic pop-up and the push-up lever pushed the toast up only about an inch. Barely sufficient for a full-sized piece of bread, and completely insufficient for a bagel or English muffin; you had to reach down in the toaster with a fork (or turn it upside down) to get those out. And no, that is _not_ safe.
I realize this is all completely off-topic...
I'm sure he meant that the robot overlords of 2026 would be using next-gen phones and computers. Built-ins.
"you have suddenly introduced a whole host of new variables" How about a controlled experiment? Send two identical machines up in a single launch; turn one of them on but not the other. If they begin separating in the expected direction, you have something. You could even switch the direction the running machine is pointed in, switch which one is turned on, etc.
I of course agree that we would want to find out how it works, if for no other reason than the fact that we might have built a very inefficient machine. But I think there's more than one way to do the science, and a controlled experiment out in space seems one perfectly reasonable way to do it.
Right, a huge diff--because in order to accelerate using propellant, at each instant you have to accelerate the propellant that you haven't used yet as well as the engines, power generating device, and the cargo. (You could use the propellant all at once, to some approximation, but the resulting acceleration would pretty much squish any cargo, human or otherwise.) And the propellant is a huge mass in comparison to the rest. (Mass of fully fueled Saturn V, 3 million kilos; mass of payload to lunar orbit: 50 thousand kilos.) With a propellant-less drive, you only need to accelerate the engines, power generating device, and cargo.
Now if the theory of how this drive works is correct, is there any way to shield the ship--or at least part of it (presumably not the engine) from the Unruh radiation? Thereby reducing its mass...
Don't bash this idea
> There are 12 inches in a foot.
Yeah, but there are only 10 toes on my foot.
You may be on to something: 73 is the sixth emirp (as mentioned in the OP). But even more significantly, 73 = 42 + 31, where 31 is the third emirp.
I'm sure that's significant...
BTW, I wrote the above--stupidly, I wasn't logged in.
So that's what Imperial Storm Troopers wear!
The headline is misleading; from the article: "There are currently no plans in the pipeline to clone and produce humans in a bid to eradicate disease, but Xiaochun has said that this can change if people become more open to the idea of it." Time for a Sixth Day law.
BTW, there's a possibly more reputable article (from Dec 2015, but basically same content) here: http://phys.org/news/2015-12-c...
IMHO, this story (I'll assume it's true, and not just boasting or wishful thinking) deserves to be told in a much wider forum (like the Wall Street Journal), with the X, Y and Z replaced by real names, and especially with your company's name.