What are the "basic things it can't do"? (I'm not disagreeing, I'm just asking what you mean.)
BTW, I would not say that real-time audio translation is a solved problem. It only works for a handful of well-resourced languages in restricted domains, where S2T and MT (and to a lesser extent T2S) work reasonably well.
They cite him in the article, so they're aware of him. I haven't read the article enough to know how this relates, but I think they're proposing another way to infer causation from causality, beyond the structural modeling (and experimental methods) that Pearl describes.
I don't know about that, but I can tell you that global warming causes the ground to get harder. Proof: when I was younger, I could camp out on the ground with no air mattress; can't do that now.
It may be obvious, but that doesn't mean it isn't contested. An example (which Pearl uses in his book Causality) is lung cancer and smoking. It was obvious to most people that smoking caused lung cancer, but another possibility was that there was a genetic predisposition to lung cancer, and that genetic factor also caused people to want to smoke. The tobacco industry in fact argued this, and (IIUC) it took some time before the direction of causation could be established in the legal sense.
An example I heard about just yesterday involved exercise and health. The question was not so much whether exercise improved health (that is obvious), but how the causation worked. The study I read about said exercise had been shown to cause methylation of DNA. Establishing that causal relationship was done experimentally by having people exercise one leg and not the other.
Causation in economics is also hard to establish (I'm told--I'm glad not to be an economist).
My first guess (I don't claim any expertise) is that a visual recognition system by itself, without some kind of world model, is doomed to fail.
Putting it differently, I think I would not do well on recognizing arbitrary sorts of images if I didn't have some idea what those images corresponded to in the real world. Photomicrographs of structures I don't know about from experience, for example.
On the other hand, numerals don't correspond to anything in the real world, and the linked-to website had examples where the vision system seemed to be hallucinating numerals in static. I don't think I can claim to have a model of what numerals are in the real world.
But we can recognize an object in a photograph or drawing, too, and there the differences in the stream of slightly different images must surely be noise, right?
Right--the meaning of the term "language" (as opposed to "dialect", "written language", etc.) is inherently fuzzy. That said, Mandarin and the other varieties of Chinese are probably more or less mutually unintelligible, in the sense that if you don't get early exposure to both varieties, or you grow up with a non-Mandarin variety and learn Mandarin in the classroom, then you won't understand the other variety. I'm told that most Chinese people can read Mandarin, but that comes from classroom instruction. And as you note, ZHO is a cover term in ISO 639-2 and 639-3 for the macro-language "Chinese", which is intended to encompass a number of varieties, but makes no claim that those varieties are mutually intelligible.
The same situation arises with a number of other "languages", such as Arabic. Indeed, there are varieties of English which speakers of most other varieties don't understand without sufficient exposure, although they are all written more or less the same.
Not sure what you mean by "multi-generational translation", I assume you're not referring to the human generations since Christ. But you also can't be referring to Hebrew/Greek --> English-->Mandarin, since the Chinese translations in use were translated with reference to the original languages (as most Bible translations into other languages are).
There are lots of things I could say, but I'll just say one: How did you measure information density?
Since I doubt that you actually measured the information density, here are two ways you could do that: 1) Look at page counts of translations of English books from/to Spanish. I don't think they'll be very different. 2) Or if you want to go for the spoken language, look at audio books.
Taking off my linguistic hat (which tells me that Spanish is just a daughter of Latin, neither better nor worse), I could say that Spanish is an improved version of Latin.
Thank you, this is the most cogent-looking response I've seen here. (I say cogent-*looking* because I'm not in a position to evaluate it; but you sound like you're at least rational:-).)
If I were a robot, able to exist in a vacuum and needing no energy besides some self-contained nuclear power plant and/or (stellar) light, I don't think I'd be too interested in earth-like planets. It's much safer out in space: no rust or dust to speak of, no competing biological organisms, no volcanoes or planet-quakes, no cliffs to fall off of, no trees or rocks getting in my way when I want to go somewhere, no wind resistance, and no wars on planets where there is intelligent life. So if these robots exist, I would think they would just steer clear of inhabited planets, and maybe of all planets.
He's not saying "therefore they did it", he's responding to Doriany's contention that the attack against Sony "achieve nothing of any value" for the North Koreans. In other words, Frosty Piss is not claiming this is evidence of guilt, but he is saying that it's a possible motive.
> Why would North Korea reveal its capabilities and tactics
They didn't reveal it, the FBI did.
> to achieve nothing of any value.
If you think that, then you don't know who Kim Jong-Un is. Stopping this movie is an obvious goal for NK; they complained about it in the UN months ago. That's not proof that they did it (or hired someone to do it), but saying retaliating against Sony is not of value to them displays a lack of cultural understanding, to say the least.
If you want an analogy, you might look at the "Jyllands-Posten Muhammad cartoons controversy".
I did too, until they ruined their user interface several years ago. And in the face of incredible opposition from the users (tens of thousands of negative posts, not a single positive post that I ever saw), they stuck with the crummy interface.
Maybe you could take 1/40th of a Zirconium atom and make a transistor out of it. I don't know how to do it, but at least in principle it should be possible. Now taking 1/40th of a Hydrogen or Helium atom, that would be difficult...
Slashdot Mirror
I believe Mark Twain explained that: http://www.crossmyt.com/hc/lin...
Huh? Someone said that, or what?
What are the "basic things it can't do"? (I'm not disagreeing, I'm just asking what you mean.)
BTW, I would not say that real-time audio translation is a solved problem. It only works for a handful of well-resourced languages in restricted domains, where S2T and MT (and to a lesser extent T2S) work reasonably well.
They cite him in the article, so they're aware of him. I haven't read the article enough to know how this relates, but I think they're proposing another way to infer causation from causality, beyond the structural modeling (and experimental methods) that Pearl describes.
I don't know about that, but I can tell you that global warming causes the ground to get harder. Proof: when I was younger, I could camp out on the ground with no air mattress; can't do that now.
Now, now. (https://www.youtube.com/watch?v=Hw5OoVXeJbU)
It may be obvious, but that doesn't mean it isn't contested. An example (which Pearl uses in his book Causality) is lung cancer and smoking. It was obvious to most people that smoking caused lung cancer, but another possibility was that there was a genetic predisposition to lung cancer, and that genetic factor also caused people to want to smoke. The tobacco industry in fact argued this, and (IIUC) it took some time before the direction of causation could be established in the legal sense.
An example I heard about just yesterday involved exercise and health. The question was not so much whether exercise improved health (that is obvious), but how the causation worked. The study I read about said exercise had been shown to cause methylation of DNA. Establishing that causal relationship was done experimentally by having people exercise one leg and not the other.
Causation in economics is also hard to establish (I'm told--I'm glad not to be an economist).
So no, I don't think causation is always obvious.
My first guess (I don't claim any expertise) is that a visual recognition system by itself, without some kind of world model, is doomed to fail.
Putting it differently, I think I would not do well on recognizing arbitrary sorts of images if I didn't have some idea what those images corresponded to in the real world. Photomicrographs of structures I don't know about from experience, for example.
On the other hand, numerals don't correspond to anything in the real world, and the linked-to website had examples where the vision system seemed to be hallucinating numerals in static. I don't think I can claim to have a model of what numerals are in the real world.
I guess if your dictionary said "(to) thrown" is a verb, then yes, you ought to throw it out.
Ceci n'est pas une pipe.
But we can recognize an object in a photograph or drawing, too, and there the differences in the stream of slightly different images must surely be noise, right?
Right--the meaning of the term "language" (as opposed to "dialect", "written language", etc.) is inherently fuzzy. That said, Mandarin and the other varieties of Chinese are probably more or less mutually unintelligible, in the sense that if you don't get early exposure to both varieties, or you grow up with a non-Mandarin variety and learn Mandarin in the classroom, then you won't understand the other variety. I'm told that most Chinese people can read Mandarin, but that comes from classroom instruction. And as you note, ZHO is a cover term in ISO 639-2 and 639-3 for the macro-language "Chinese", which is intended to encompass a number of varieties, but makes no claim that those varieties are mutually intelligible.
The same situation arises with a number of other "languages", such as Arabic. Indeed, there are varieties of English which speakers of most other varieties don't understand without sufficient exposure, although they are all written more or less the same.
Not sure what you mean by "multi-generational translation", I assume you're not referring to the human generations since Christ. But you also can't be referring to Hebrew/Greek --> English-->Mandarin, since the Chinese translations in use were translated with reference to the original languages (as most Bible translations into other languages are).
Syntax. No dog understands syntax, not one.
There are lots of things I could say, but I'll just say one: How did you measure information density?
Since I doubt that you actually measured the information density, here are two ways you could do that:
1) Look at page counts of translations of English books from/to Spanish. I don't think they'll be very different.
2) Or if you want to go for the spoken language, look at audio books.
Taking off my linguistic hat (which tells me that Spanish is just a daughter of Latin, neither better nor worse), I could say that Spanish is an improved version of Latin.
Thank you, this is the most cogent-looking response I've seen here. (I say cogent-*looking* because I'm not in a position to evaluate it; but you sound like you're at least rational :-).)
If I were a robot, able to exist in a vacuum and needing no energy besides some self-contained nuclear power plant and/or (stellar) light, I don't think I'd be too interested in earth-like planets. It's much safer out in space: no rust or dust to speak of, no competing biological organisms, no volcanoes or planet-quakes, no cliffs to fall off of, no trees or rocks getting in my way when I want to go somewhere, no wind resistance, and no wars on planets where there is intelligent life. So if these robots exist, I would think they would just steer clear of inhabited planets, and maybe of all planets.
Bureau 121.
North Korea has resources; it's a question of how they deploy them. Hint: it's not to the ordinary citizenry.
He's not saying "therefore they did it", he's responding to Doriany's contention that the attack against Sony "achieve nothing of any value" for the North Koreans. In other words, Frosty Piss is not claiming this is evidence of guilt, but he is saying that it's a possible motive.
> Why would North Korea reveal its capabilities and tactics
They didn't reveal it, the FBI did.
> to achieve nothing of any value.
If you think that, then you don't know who Kim Jong-Un is. Stopping this movie is an obvious goal for NK; they complained about it in the UN months ago. That's not proof that they did it (or hired someone to do it), but saying retaliating against Sony is not of value to them displays a lack of cultural understanding, to say the least.
If you want an analogy, you might look at the "Jyllands-Posten Muhammad cartoons controversy".
I did too, until they ruined their user interface several years ago. And in the face of incredible opposition from the users (tens of thousands of negative posts, not a single positive post that I ever saw), they stuck with the crummy interface.
"we discovered that time comes in discrete quantum units": we did? Citation?
Hey, I can do three decimal places with my slide rule. At least toward the left-hand end.
Maybe you could take 1/40th of a Zirconium atom and make a transistor out of it. I don't know how to do it, but at least in principle it should be possible. Now taking 1/40th of a Hydrogen or Helium atom, that would be difficult...