If you're dubious about "special compilation", then use GCC to compile clang. (It's not sufficient, but using a different compiler to do the compilation of the compiler is a necessary step.)
You don't understand the problem. All strains of banana are monocultures, because their seeds never ripen. This means that you can only reproduce them by cloning (an easy process for bananas and I think dating back to per-columbian times.).
Now even clones accumulate mutations, and that's why there are several different strains of banana, but there's no sexual recombination of genes. So you end up with a monoculture in each area, with each area having a slightly different accumulation of mutations. But they are quite similar. Similar enough that a fungus can wipe out an entire crop leaving no survivors (because there's no useful amount of genetic variation). And because the fungus is not a monoculture, it can more readily adapt to various different strains. Thus bananas varieties are inherently subject to being wiped out by a fungus, and there's a fungus that's been doing this since the 1930's (at least). Rapid long distance transportation makes things worse, as it allows the fungus to spread more rapidly and widely.
And this problem exists for ALL strains of banana. There are some for which the fungus has not yet adapted, but their time will come.
Bananas are one crop for with GMO techniques are necessary. Naval Oranges will be another. (IIUC, seedless watermelons are produced by chemically treating the seeds that will grow into the watermelon plant, so this is a different matter.)
Do note, however, that while bananas and naval oranges are extreme cases, the problem exists to a lesser extent in all monoculture crops.
It's not simply because we can't afford to say no. The costs started rising when the required paperwork started rising. In all history before the 1950's a doctor's office was usually the doctor, or the doctor and his wife. Sometimes a doctor and a nurse. Now (in the US) there's a staff to handle the paperwork, and a single doctor can't afford to practice. First there was government paperwork, then HMOs started demanding more, and there's government, HMOs, insurance companies, etc.
The stupid thing is that computers, which should have freed us from paperwork, were instead conscripted to drastically increase the amount of paperwork (although now it often never involved paper per se).
I'm rather sure you are overestimating the intelligence of Alex and his relatives, and of the corvids. Yes, they're quite intelligent for birds, and Alex demonstrated minimal understanding of grammar. This doesn't make them even nearly as smart as a dog.
OTOH, as I've asserted before, intelligence is not a thing, but rather a collections of capabilities. Some of the capabilities include the capability to use other capabilities as tools to accomplish some limited tasks. And I deny the existence of this "general intelligence" thing which IQ tests claim to measure. What they actually measure is a collection of socially approved capabilities. So, for example, the correlation between having a high IQ and being a successful musician or author or politician is, if anything, slightly negative. Those careers require capabilities which are not measured by IQ tests and which appear to vary in a way independent of the capabilities measured by IQ tests. (Yeah, a certain minimal level of the capabilities measured by IQ tests is needed, but that's possessed by almost anyone with a decent social background who doesn't have something physically wrong with them. And more than the minimum isn't that helpful in those professions.)
There is an indication that this may not be possible without extensive interventions...and I'm thinking of a nanobot immune system.
Indications are not, of course, proof. The indication was the number of immune stem cell variations found in a 120+ year old human. It had, IIRC, been reduced to below 10. (I want to say reduced to a single variant, but I don't really remember that.) Now I suppose that you might get around this by banking umbilical cord blood, and using that as a source of immune cell variation, but these new immune cells wouldn't remember the diseases you've already been exposed to and learned to deal with.
So. It's my expectation that to significantly extend the maximum lifespan of a human would require an artificial replacement immune system, and while dogs probably have more basic capabilities in that direction, they would probably need the same replacement to make it much beyond 120.
FWIW, I haven't yet seen evidence that Google sells consumer data. I'll admit they might, but their business model revolves around being the keeper of the data and acting as a man in the middle whenever anyone wants to use the data. I.e., they sell access to the individuals, not who the individuals are. Once they sell who the people are they loose their monopoly.
There's also the lamentable ease with with anonymized data has been repeatedly de-anonymized.
MAYBE they actually did it this time. Perhaps. I guess. But there have been so many failures, often apparently without malice on the part of the anonymizing agent, that I find it difficult to believe.
C also embeds well into Ruby, C++, Ada, and D. I'm not sure about other languages, but many LISP and Scheme dialects seem to claim to work well with C. OTOH, I've never seen a Snobol interface to C.
Why should it? I notice that the Linux version is speced to run on an updated Ubuntu LTS. Doesn't claim to run on anything else. Probably they tested it and it happened to work, or someone noticed it could do so with minimal tweaking. Otherwise they'd probably have open source an Apple only version. And it would still be legitimately open source. Open source doesn't guarantee cross platform.
If your are worried that it may be fearful and attack us, then you are anthropomorphizing it invalidly. This doesn't mean that in optimally pursuing it's goals it wouldn't undertake actions that in a human would be fear driven, but in a well-understood goal system this would more properly be called constructing sub-goals to optimize pursuit of it's major goals. E.g., you can't turn the universe into paper clips is some intrusive individual insists on turning you off.
This makes design of the major goals a real problem, but it also means that it won't alter its major goals or even desire to do so. It may, however, desire to prevent YOU from altering them.
The mechanism is not a complete mystery. We know, e.g., that it involves imagining yourself in the environment and solving the problem. That it doesn't need to be visual, but in humans it usually is. However kinesthetic modeling works just as well if your sensorium is set up that way. That it depends on having a large library of experiences that are similar in some way to the action or result being predicted. Etc. Some of this is because of experimental brain surgery. Some is derived from lab animals. Some is due to modern brain scanning tools. Etc. Often the resolution of the resulting knowledge leaves a lot to be desired, and this makes it difficult to produce working models, but that's still a lot more than "a complete mystery". 20-30 years ago it might have been defensible to term it "a complete mystery", but even then it would have been dubious.
Modern chess playing programs are not machine intelligence. At least they weren't 20 years ago when I studied them. They are a specialized tool that was completely understood (by someone). Calling that intelligence is a mistake. If you understand precisely how something operates, then it's not intelligence, it's an algorithm, template, or something like that.
There are modern programs which are intelligent. It's appropriate to say this because even though to source code is available, nobody understands how they work to achieve correct answers. Many of these evolved through some variation of genetic programming, or have embedded modules that did. It would be a waste to use them for playing chess, though you could do so. You'd need to train them for a long time, and they would still often not be as capable as a specialized non-intelligent program. OTOH, if you wanted a general game playing program, intelligence would be the way to go. Then you could teach it to play a new game rather than having the game essentially built into the program.
I believe that the idea that "general intelligence" exists is a mistake. There are just sets of tools that can do particular jobs. This looks like intelligence to us when we don't look closely enough at some particular tool to see how it works. A few of the "tools" that people have are high level tools that let them use "black box" library routines without understanding them. We don't solve differential equations to catch a ball, we invoke a built-in tool-script that has had lots of adjustable parameters tweaked to work in our body. A similar action happens when a human plays chess. We don't actually use alpha-beta pruning, but we've got a built-in tool that has about the same effect....but which is a lot more adjustable. Et multitudinous cetera.
As I understand it (see other posts for the official line sold to the public) the purpose is to deprive people of the ability to hire lawyers to defend themselves. I don't believe that there is any requirement that guilt be proven, or even that charges be formally filed. The first time I heard of it being used it was against a doctor who was prescribing more pain medication than the DIA thought was appropriate.
Yeah, you're right. But I tend to believe this claim anyway. Some of the things China is blamed for they actually do.
OTOH, it's not clear that this was an action by the Chinese government. (The summary didn't say that was even claimed.) And if it were, it's not clear that it would be the policy of the government rather than some loose cannon. (I assume they've got just as many as we do.)
The reported response, however, seems more PR than anything else. (Again, just based on the summary.) This shouldn't be surprising. I bet China's government is even more labyrinthine than ours.
Who supposedly owns the computers? If the school buys them and issues them to the students, then the argument if invalid, but if the students (families) must purchase the computers, then it is a valid argument.
Similarly, if in the corporate environment the company provides the equipment, then the company has total right to say what's installed, but if the employee provides the equipment they have no such right. (Power, yest, but not right. The same goes for schools...except that schools are, generally, an arm of the government.)
Price is one thing. Suitability is another. I'm willing to accept that SSDs are excellent cache storage, and in that application their increased speed justifies the larger price. But what I'm looking for is durable backup capability, and so far SSDs don't seem to fit the picture even at half or less the price of usb HDs. Their failure is to sudden and too complete. If DVDs weren't so small I wouldn't even be looking at them.
FWIW, I was told at one point that SSDs tend to loose their contents over time due to charge leak. If so I doubt that they'll ever be suitable for backup storage. And the smaller the circuit, the worse the leakage tends to be, so I don't expect new generations to be improved in this respect. (It would sure be nice if I were wrong.)
That hasn't been my experience. I'm considering them for backup storage, and I've had SSD's die without warning, apparently from being removed from power for too long. Not acceptable for a backup.
OTOH, what I'm talking about were thumb drives. But why should I think other removable SSDs would be different?
People claim all sorts of things. While there are lots of problems with a space elevator on a world as large as Earth, energy efficiency isn't one of them.
Personally, I doubt that a space elevator will ever be practical on Earth, but it should be on Mars, and it definitely would be on the Moon. For Earth I'd favor something like the pinwheel. You can think of the pinwheel as a rotating space elevator that doesn't reach as far down as the ground. (You'd probably want to not reach further down than the upper stratosphere to minimize frictional losses.) You fly up to meet the descending arm at the bottom, and unload cargo onto it. Descending cargo can be handled the same way, or you could use a combination of parachutes and lifting bodies. You need to balance freight going up and coming down or you get orbital decay...either it lifts too high or comes down too far, but this can be handled by a station keeping ion rocket, possibly of Vasimir design. Reaching a height is a lot cheaper than going into orbit. This does require a large orbital mass.
Slashdot Mirror
If you're dubious about "special compilation", then use GCC to compile clang. (It's not sufficient, but using a different compiler to do the compilation of the compiler is a necessary step.)
You don't understand the problem. All strains of banana are monocultures, because their seeds never ripen. This means that you can only reproduce them by cloning (an easy process for bananas and I think dating back to per-columbian times.).
Now even clones accumulate mutations, and that's why there are several different strains of banana, but there's no sexual recombination of genes. So you end up with a monoculture in each area, with each area having a slightly different accumulation of mutations. But they are quite similar. Similar enough that a fungus can wipe out an entire crop leaving no survivors (because there's no useful amount of genetic variation). And because the fungus is not a monoculture, it can more readily adapt to various different strains. Thus bananas varieties are inherently subject to being wiped out by a fungus, and there's a fungus that's been doing this since the 1930's (at least). Rapid long distance transportation makes things worse, as it allows the fungus to spread more rapidly and widely.
And this problem exists for ALL strains of banana. There are some for which the fungus has not yet adapted, but their time will come.
Bananas are one crop for with GMO techniques are necessary. Naval Oranges will be another. (IIUC, seedless watermelons are produced by chemically treating the seeds that will grow into the watermelon plant, so this is a different matter.)
Do note, however, that while bananas and naval oranges are extreme cases, the problem exists to a lesser extent in all monoculture crops.
It's not simply because we can't afford to say no. The costs started rising when the required paperwork started rising. In all history before the 1950's a doctor's office was usually the doctor, or the doctor and his wife. Sometimes a doctor and a nurse. Now (in the US) there's a staff to handle the paperwork, and a single doctor can't afford to practice. First there was government paperwork, then HMOs started demanding more, and there's government, HMOs, insurance companies, etc.
The stupid thing is that computers, which should have freed us from paperwork, were instead conscripted to drastically increase the amount of paperwork (although now it often never involved paper per se).
I'm rather sure you are overestimating the intelligence of Alex and his relatives, and of the corvids. Yes, they're quite intelligent for birds, and Alex demonstrated minimal understanding of grammar. This doesn't make them even nearly as smart as a dog.
OTOH, as I've asserted before, intelligence is not a thing, but rather a collections of capabilities. Some of the capabilities include the capability to use other capabilities as tools to accomplish some limited tasks. And I deny the existence of this "general intelligence" thing which IQ tests claim to measure. What they actually measure is a collection of socially approved capabilities. So, for example, the correlation between having a high IQ and being a successful musician or author or politician is, if anything, slightly negative. Those careers require capabilities which are not measured by IQ tests and which appear to vary in a way independent of the capabilities measured by IQ tests. (Yeah, a certain minimal level of the capabilities measured by IQ tests is needed, but that's possessed by almost anyone with a decent social background who doesn't have something physically wrong with them. And more than the minimum isn't that helpful in those professions.)
There is an indication that this may not be possible without extensive interventions...and I'm thinking of a nanobot immune system.
Indications are not, of course, proof. The indication was the number of immune stem cell variations found in a 120+ year old human. It had, IIRC, been reduced to below 10. (I want to say reduced to a single variant, but I don't really remember that.) Now I suppose that you might get around this by banking umbilical cord blood, and using that as a source of immune cell variation, but these new immune cells wouldn't remember the diseases you've already been exposed to and learned to deal with.
So. It's my expectation that to significantly extend the maximum lifespan of a human would require an artificial replacement immune system, and while dogs probably have more basic capabilities in that direction, they would probably need the same replacement to make it much beyond 120.
Define your terms. How, for instance, do you count running out into traffic?
FWIW, I haven't yet seen evidence that Google sells consumer data. I'll admit they might, but their business model revolves around being the keeper of the data and acting as a man in the middle whenever anyone wants to use the data. I.e., they sell access to the individuals, not who the individuals are. Once they sell who the people are they loose their monopoly.
There's also the lamentable ease with with anonymized data has been repeatedly de-anonymized.
MAYBE they actually did it this time. Perhaps. I guess. But there have been so many failures, often apparently without malice on the part of the anonymizing agent, that I find it difficult to believe.
Apple may be more serious about it. After all, they are mainly a hardware company.
C also embeds well into Ruby, C++, Ada, and D. I'm not sure about other languages, but many LISP and Scheme dialects seem to claim to work well with C. OTOH, I've never seen a Snobol interface to C.
Why should it? I notice that the Linux version is speced to run on an updated Ubuntu LTS. Doesn't claim to run on anything else. Probably they tested it and it happened to work, or someone noticed it could do so with minimal tweaking. Otherwise they'd probably have open source an Apple only version. And it would still be legitimately open source. Open source doesn't guarantee cross platform.
If your are worried that it may be fearful and attack us, then you are anthropomorphizing it invalidly. This doesn't mean that in optimally pursuing it's goals it wouldn't undertake actions that in a human would be fear driven, but in a well-understood goal system this would more properly be called constructing sub-goals to optimize pursuit of it's major goals. E.g., you can't turn the universe into paper clips is some intrusive individual insists on turning you off.
This makes design of the major goals a real problem, but it also means that it won't alter its major goals or even desire to do so. It may, however, desire to prevent YOU from altering them.
The mechanism is not a complete mystery. We know, e.g., that it involves imagining yourself in the environment and solving the problem. That it doesn't need to be visual, but in humans it usually is. However kinesthetic modeling works just as well if your sensorium is set up that way. That it depends on having a large library of experiences that are similar in some way to the action or result being predicted. Etc. Some of this is because of experimental brain surgery. Some is derived from lab animals. Some is due to modern brain scanning tools. Etc. Often the resolution of the resulting knowledge leaves a lot to be desired, and this makes it difficult to produce working models, but that's still a lot more than "a complete mystery". 20-30 years ago it might have been defensible to term it "a complete mystery", but even then it would have been dubious.
Modern chess playing programs are not machine intelligence. At least they weren't 20 years ago when I studied them. They are a specialized tool that was completely understood (by someone). Calling that intelligence is a mistake. If you understand precisely how something operates, then it's not intelligence, it's an algorithm, template, or something like that.
There are modern programs which are intelligent. It's appropriate to say this because even though to source code is available, nobody understands how they work to achieve correct answers. Many of these evolved through some variation of genetic programming, or have embedded modules that did. It would be a waste to use them for playing chess, though you could do so. You'd need to train them for a long time, and they would still often not be as capable as a specialized non-intelligent program. OTOH, if you wanted a general game playing program, intelligence would be the way to go. Then you could teach it to play a new game rather than having the game essentially built into the program.
I believe that the idea that "general intelligence" exists is a mistake. There are just sets of tools that can do particular jobs. This looks like intelligence to us when we don't look closely enough at some particular tool to see how it works. A few of the "tools" that people have are high level tools that let them use "black box" library routines without understanding them. We don't solve differential equations to catch a ball, we invoke a built-in tool-script that has had lots of adjustable parameters tweaked to work in our body. A similar action happens when a human plays chess. We don't actually use alpha-beta pruning, but we've got a built-in tool that has about the same effect....but which is a lot more adjustable. Et multitudinous cetera.
Yeah, that's the way I read it. Mind you, it's gotten a lot worse than what I suppose the original purpose to have been.
As I understand it (see other posts for the official line sold to the public) the purpose is to deprive people of the ability to hire lawyers to defend themselves. I don't believe that there is any requirement that guilt be proven, or even that charges be formally filed. The first time I heard of it being used it was against a doctor who was prescribing more pain medication than the DIA thought was appropriate.
Well...Nixon did open diplomatic relations with "Red China", but I guess compared to the current crop of Republicans he was a plaster saint.
Yeah, you're right. But I tend to believe this claim anyway. Some of the things China is blamed for they actually do.
OTOH, it's not clear that this was an action by the Chinese government. (The summary didn't say that was even claimed.) And if it were, it's not clear that it would be the policy of the government rather than some loose cannon. (I assume they've got just as many as we do.)
The reported response, however, seems more PR than anything else. (Again, just based on the summary.) This shouldn't be surprising. I bet China's government is even more labyrinthine than ours.
Who supposedly owns the computers? If the school buys them and issues them to the students, then the argument if invalid, but if the students (families) must purchase the computers, then it is a valid argument.
Similarly, if in the corporate environment the company provides the equipment, then the company has total right to say what's installed, but if the employee provides the equipment they have no such right. (Power, yest, but not right. The same goes for schools...except that schools are, generally, an arm of the government.)
Price is one thing. Suitability is another. I'm willing to accept that SSDs are excellent cache storage, and in that application their increased speed justifies the larger price. But what I'm looking for is durable backup capability, and so far SSDs don't seem to fit the picture even at half or less the price of usb HDs. Their failure is to sudden and too complete. If DVDs weren't so small I wouldn't even be looking at them.
FWIW, I was told at one point that SSDs tend to loose their contents over time due to charge leak. If so I doubt that they'll ever be suitable for backup storage. And the smaller the circuit, the worse the leakage tends to be, so I don't expect new generations to be improved in this respect. (It would sure be nice if I were wrong.)
That hasn't been my experience. I'm considering them for backup storage, and I've had SSD's die without warning, apparently from being removed from power for too long. Not acceptable for a backup.
OTOH, what I'm talking about were thumb drives. But why should I think other removable SSDs would be different?
Give proprietary icons codes above 2^24 and I'm ok with that. Of course, no fonts will implement them, but that's ok with me too.
People claim all sorts of things. While there are lots of problems with a space elevator on a world as large as Earth, energy efficiency isn't one of them.
Personally, I doubt that a space elevator will ever be practical on Earth, but it should be on Mars, and it definitely would be on the Moon. For Earth I'd favor something like the pinwheel. You can think of the pinwheel as a rotating space elevator that doesn't reach as far down as the ground. (You'd probably want to not reach further down than the upper stratosphere to minimize frictional losses.) You fly up to meet the descending arm at the bottom, and unload cargo onto it. Descending cargo can be handled the same way, or you could use a combination of parachutes and lifting bodies. You need to balance freight going up and coming down or you get orbital decay...either it lifts too high or comes down too far, but this can be handled by a station keeping ion rocket, possibly of Vasimir design. Reaching a height is a lot cheaper than going into orbit. This does require a large orbital mass.
Please do not set up the EFF for regulatory capture.