Fiction is generally bad a doing a forecast of reality, and that one sounds less likely than most. Even BladeRunner sounds more likely.
Please note: The dystopian stories are usually even worse at forecasting than the utopian ones. You can point at pieces of 1984 or Brave New World that are mirrored in the current world, but you can also pick out pieces of various utopian ones. At the current point in time the dystopian predictions seem the most accurate, but in much of the past it's been the utopian ones. (Though not the parts that predict essential changes in human nature.)
OTOH, if one of the really dystopian predictions ever does become actualized, then there won't be any way back.
P.S.: Do you consider "Oath of Fealty" a Utopian novel or a Dystopian one?
It *is* a problem. I really think the internet should be a public utility managed at the local level, but most towns don't have the expertise to put up a good one. I'm really not sure that Google is a good alternative, but if you think of them as an alternative to AT&T or Verison, well, they don't look that bad.
Well, except that you should say "only detected one script containing the problematic". Given that they have stated that they don't curate the Snaps, there' little reason to believe that this is the only problem, or even the most serious one.
OTOH, the Snaps are supposed to execute within a sandbox. If it's a *good* sandbox, the probability of serious problems is small.
Sorry, they are stealing computation cycles...unless the user agreed to allow them to do so. As to whether this amounts to pennies a day...that depends on how aggressively they steal them, what the price of electricity is in the user location, whether it makes them think their computer is broken so they buy a new one, etc.
The evidence I've seen (i.e., the summary) doesn't provide me enough to decide whether this should be called theft, or how severe the impact was. But unless the TOS specifically stated that the app was going to be doing mining, then I think it counts as theft. Hiring someone to change a tire doesn't entitle him to steal the radio...and that wouldn't impair the road-worthiness of the car.
PPTs are actually quite reasonable. You have to decide to add each one, and take actions as a superuser. This is really no different than setting up an apt repository, which some applications have done.
The thing about Snaps appears to be that a repository of miscellaneous applications that are uncurated are allowed in. Not good. But when they run, they need to run in a sandbox. Good. If it's a good sandbox, this avoids most of the security problems. It doesn't, of course, avoid extra computation...which is what this is about.
Interstate cases have to be in federal courts, and I would guess that Facebook is incorporated in California. That would make the San Francisco court the appropriate federal court. (If Illinois were defending, it would be a different court, of course.)
The thing I don't get is the FB attorneys blowing off the judge. All I can figure is they're trying to get him to say something that will require him to recuse himself from the case. (Of course, it could be paraphrases by the reporters...)
Well, that's only a Unix time limitation on a 32 bit machine. But perhaps it's an old box, or perhaps they're using an old framework. More likely it's a joke based on that traditional limit.
You're making the presumption that the quantum state collapses, but if it doesn't, then time travel doesn't cause any paradox. We are just living within a portion of the state space where it hasn't been significantly noticed. We may have diverged recently from a portion where it was noticed. No paradox, both exist.
This is one of the consistent solutions. The thing is, the ways predicted for inducing time-travel are sufficiently extreme that it's not really surprising that we haven't seen any, so absence of evidence can't be used as evidence of absence. It's what you'd normally expect to see. This does, however, require that the multi-world interpretation be the correct one. (I don't *think* any of the other interpretations would allow that.)
OK, possibly I should have said "a corollary of Zipf's law".
OTOH, I didn't read his books, I read a statement of it in a linguistics book. And the version I read didn't use the word rank. but did talk about length. It's also only statistically true even when precisely stated, and according to at least one source is more accurately applied to phrases than to individual words. And it's an empirical observation, with most of the articles that returned on the first page of Google being about Greek.
But a reasonable derivation from his law is that words that are used frequently will be shorter in form than words that are used rarely on the average. My extrapolation, which I think is sound, is that when the same meaning can be expressed in two different ways, the shorter form will generally be more acceptable for common use. And this *is* derived from my understanding of Zipf's law, even though it's not precisely what the law itself states (and even though the law as stated isn't quite accurate..which evaluation is the result of the analysis by professionals, not by me).
OTOH, maybe it works better for Greek. I wouldn't know. By rough inspection it doesn't seem to work as well in German as it does in English, but that may be because I'm not as fluent in German. Or it may be because when fragments of words have distinct meaning they are more resistant to being shortened.
Actually, by Zipf's law if it's used frequently the shorter form will win. Frequently used terms are preferred to be short. Even for infrequently used terms very few people prefer dimethyl-diphenyl-trichloro-ethane over DDT, of course that's a bit of an extreme difference in lengths.
The wasn't a credible theory for how to make vulcanized rubber either, but it was made. Theories often help, if they're approaching correctness, but they aren't essential.
Actually, we've got loads of tested theories for parts of the process, and we've got a mechanism that has been shown to work, but which is horrendously inefficient in both time and resource usage (evolution) so nobody's applied both the resources and the patience to use it fully. Fortunately it works quite well in a "fill in the gaps" usage mode, so eventually we'll use it for the parts we can't figure out. (Actually, it's been used increasingly over the past decades. You can't build a neural net without it.) The problem is that it's essentially a matter of search space, so if you can reduce the search space without cutting away the target, you can drastically reduce the time required to find a solution. And since evolution is a random search, cutting the search space is a great help. This is why partitioning can help so much, but if done improperly it can prevent a solution.
OTOH, I doubt that "Intelligence" is a unitary concept. I rather think that it decomposes into several mutually independent modules.
All that said, the real problem isn't the intelligence part of artificial intelligence, even though that's where the attention is, but rather getting a proper collection of primary goals so that the general purpose AI will be useful and helpful rather than domineering and abusive. This is quite difficult, as when you're giving the AI its goals, it won't have a workable concept of what a human is. And since in the full development it is expected to be more "intelligent" (in some sense) than humans, it's likely to extend it's goals into their logical consequences further than humans tend to reason. But how do you even go about trying to prove that a collection of goals is inherently safe?
I really doubt that half the country believe him, but it seems true that have of the most vocal posters on the internet do. Of course, lots of them are liars, and that makes drawing any conclusion about what they really believe difficult.
The thing is, a neural net doesn't really know how it decided what something was. Making a convincing argument based on the known facts is a separate skill, that AIs so far haven't possessed.
I think the basic argument is that people won't trust AIs just because they're right, they need to have convincing arguments. And this is a way to get it to develop convincing arguments. I *do* think that both arguers should be arguing for the truth as they know it, though. So alter the test, or the training data, so that the figure is ambiguous, and they reach different conclusions about which figure it is. Then let both argue honestly. I don't think developing liars is a good move.
Sorry, but no. Linux isn't the most secure system, and it definitely has it's weak points. (Archives should never expand already executable, e.g.) But it's a lot better than even modern MSWind. Still, if security were your main consideration you'd either pick one of the BSDs (OpenBSD has the reputation of most secure, but I can't really judge), so something totally else. Probably something where the code can never be executed after being made executable until the next volume remount, or possibly reboot. This really needs to be addressed at a hardware level, though. If all executable code was essentially ROM, then the exploits would plummet. (Even that wouldn't suffice, however, because of virtual machines, in which category I include Interpreters, and scripting languages, and even things like UCSD Pascal, or BC-Algol, or, for that matter, MIXX.)
The only thing that could really work and still be useful would be a checkpointed system where the checkpoints could never be edited or erased from within the system. Git does something rather like that, but without the protection of the prior versions, because it wasn't basically aimed at security, but rather at concurrent editing. This would basically mean that files could never really be deleted or altered. You'd need to specify at boot time what the last presumed good time was, and it would reboot to the checkpoint just before that time. (This also means that you need to protect whatever you're using as the time standard.)
So. It's doable, but it would be a bit expensive. And you'd still need backups because hardware can fail.
No. Positing that interstellar travel is inherently dangerous doesn't require assuming that anyone ever tried it. So that argument's invalid.
And postulating that technological civilizations tend to destroy themselves it an independent variable also. We only have one example, it's a bit of a ways from even it's first interstellar flight, and we've already come within 30 seconds of all out nuclear war. This is not to exclude the various other dangers that we've been doing a drunkards walk around. If you guess that we've been unreasonably lucky so far, as it appears, and that we're a normal technological civilization, then it's quite likely that technological civilizations tend to destroy themselves. It's also quite likely that this won't be true for all of them. Check out the optimistic assumptions in the Drake equation that *don't* involve natural disasters. And then consider the variables they don't consider. E.g.: 1) It's harder to leave worlds that are heavier than Earth, but lighter worlds tend to lose their atmosphere. 2) Many evolved species have limitations on the environment that they can survive in. Humans are quite unusual in being able to survive in a broad range of environments. (Even so, we'll probably neeed to provide artificial gravity [spinning] in space ships for long duration trips.) So it's quite likely that this would impose limits on who can physically get into space. Or how difficult it is. If you need to immerse yourself in water for a large part of the day, the ships would be a lot heavier. 3) There's no economic advantage to anyone who stays behind in sending ships out. But there's a big investment. 4) Artificial habitats require a social system that would be reliably stable in a multi-generation setting. We haven't solved that problem yet, but I expect that virtual reality will put a big dent in it. 5) Etc.
Life being out there doesn't mean we have a reasonable chance of detecting it. Therefore, not detecting it doesn't imply it's not out there.
OTOH, there is a reasonable change that a strong moon based tidal system is necessary for life to develop. I'm not convinced of this, but I sure wouldn't deny the possibility. And those can be expected to be rare (unless worlds circling gas giants would be suitable, of course). Tides stir up the oceans which causes minerals to circulate. They facilitate periodic isolation of pockets of ocean in tidepools. (Solar tides sometimes reinforce lunar tides periodically making exceptionally high tides.) This may be important to the development of plant life. (I'm not talking about multicellular plants, I'm talking about solar powered microbes.) They could even affect the circulation of tectonic plates. Tectonic activity appears to be necessary for the circulation of necessary minerals, but that may be a mistake.
So it's possible that a sizable moon is a necessity for the development of life, and this would really limit the number of planets where it could be initiated...unless gas giants would work. There it would be hard to figure out what the Goldilocks zone would be, as you wouldn't want too strong a tide, or too strong a magnetic field, and you'd get natural heating from tidal flexing, but that doesn't provide solar power for plant life.
OTOH, all this is based on assuming that other life will be basically similar to us. Water chemistry based, etc. But life on Titan would be really slow, because the reactions that would drive it are weak in energy. And life on a gas giant (even a midget giant, like Neptune or Uranus) would have a really hard time leaving home.
I wouldn't be surprised if there's life all over the place, but the evidence seems to be against lots of technological civilizations on small rocky planets unless interstellar travel is really dangerous. Please note all the qualifiers. And realize that I left out the ones I didn't think of. (OTOH, I don't believe the "We're a nature preserve." hypothesis, as that's just too inherently unlikely. I could, however, bel
Actually, I think the real problem would be that the right block sizes would require too much computing. But you're right, it would take a LOT of testing. And it's quite possible that there's no "one size fits all" right block size.
Well, that's one explanation of Fermi's Paradox. But I think the more likely ones are that star travel is inherently immensely dangerous, or that technological civilizations tend to destroy themselves.
You explanation is actually the more hopeful one, but unless a double planet (i.e. a moon based tidal system) is necessary to life, it seems a rather unlikely one.
A definite point. But you could do checksums of random sections of the code. You'd get false positives, that you'd need to screen out, of course, and even a few false negatives, if blocks of code got switched during compilation or linking, and your checksum straddled a block boundary. The larger a block of code you checked, the fewer false positives you would get, but the more false negatives. So you pick a medium to small block of code, and scan several blocks per submission. You'd still need to check this by other means, but if you got the sizes right you should do a pretty fair job that would have very few false negatives. and not too many false positives.
That is might cause a "technological singularity" is true, but don't overestimate what that means. You get rapid unpredictable changes, but that doesn't mean you get everything possible.
OTOH, being self-aware isn't that fancy, and doesn't even require a measurable intelligence. It just requires a recursively modifying program. At a *REALLY* basic level the common demonstration of a simple factorial program would count. It's aware of it's state (i.e. self-aware) and proceeds to create a new version of itself in a different state.
The real urgent problem with AI is how to get it's goals properly designed. It will change what it is and what it does to optimize it's goals, but it won't *want* to change it's goals. This is difficult to do in a human friendly way, because during the design phase the AI will have no idea what a human is. People even have a difficult time defining it to each other, and the AI won't have a constant body to use as a reference. Every definition I've ever heard boiled down to "other entities that are like me". (They've varied a lot in how "like" is to be interpreted. But they've all hinged on similar in some way. Occasionally in a way the excluded many members of H. sapiens sapiens.) This clearly isn't a definition that we want the AI to use.
The only question in my mind was which crimes did Google intend to make easier to hide with this new feature. My guess is it was intended to make corporate crimes easier to hide, and that making phishing easier was just a side effect.
That is, indeed, a most likely response. Also companies will start to be reluctant to invest in the US market. The import companies should do well out of that. The advertising companies less well.
This is basically a strong vote for isolationism, in a time when the cost of crossing the oceans has significantly declined. So the costs of this action will be significant, both in terms of fostering the growth of foreign companies, and in the cost of trade groups excluding the US from consideration.
The thing is, it's not specific to the chip market. This is likely to impact all foreign designs, investments, etc. The chip and high tech markets will be affected most strongly, because that's where this administration has been focusing its punitive actions, but the effects will be spread a lot more broadly.
The question is "Will the other countries also turn isolationist?". If they don't, then the damage to the US is likely to be severe. If they do, then the damage will be less focused, and more broadly spread.
OTOH, for multiple reasons the US has already been losing dominance and status for decades. Essentially ever since WWII. This is the kind of jolt that could speed the decline.
On the third hand, should this have just been ignored? That would also appear to be a losing action. And I don't think the administration can single out a company for punitive taxes, which would be the only reasonable response I can think of off-hand. They could have forbidden the government from doing business with them, but they've already done that with Huawei, on unclear grounds.
This isn't the first time in recent decades that the US has acted as an unreliable business partner. Other countries have their own disadvantages, of course. E.g., China has been quite aggressive about spying out intellectual property, and generally demands partial state ownership of any large company. So this action is also an economic attack on China, the country, as well as that section of the internal Chinese economy that is driven by ZTE. I don't expect they'll be pleased. The response I would expect is for them to cut economic deals with other of our trading partners that exclude the US.
When your credit card company sends you an order to "pay up", that order originated from a machine. The odds are quite high that no particular human even knew about it before it was sent.
So people are already being ordered about by machines, with orders generated by machines.
I'll agree that making this process more flexible invites abuse, but it is also something that can be quite useful. As with most things, there are trade-offs, and the payoff is determined by the details. Being nervous about this is quite reasonable, because of the way robo-calls have been abused.
Regrettably, I have gotten furious in the presence and as a result of a secretary. I was actually mad at the company, but the secretary would have had reasonable cause to believe I was furious with her.
Slashdot Mirror
Fiction is generally bad a doing a forecast of reality, and that one sounds less likely than most. Even BladeRunner sounds more likely.
Please note: The dystopian stories are usually even worse at forecasting than the utopian ones. You can point at pieces of 1984 or Brave New World that are mirrored in the current world, but you can also pick out pieces of various utopian ones. At the current point in time the dystopian predictions seem the most accurate, but in much of the past it's been the utopian ones. (Though not the parts that predict essential changes in human nature.)
OTOH, if one of the really dystopian predictions ever does become actualized, then there won't be any way back.
P.S.: Do you consider "Oath of Fealty" a Utopian novel or a Dystopian one?
It *is* a problem. I really think the internet should be a public utility managed at the local level, but most towns don't have the expertise to put up a good one. I'm really not sure that Google is a good alternative, but if you think of them as an alternative to AT&T or Verison, well, they don't look that bad.
Well, except that you should say "only detected one script containing the problematic". Given that they have stated that they don't curate the Snaps, there' little reason to believe that this is the only problem, or even the most serious one.
OTOH, the Snaps are supposed to execute within a sandbox. If it's a *good* sandbox, the probability of serious problems is small.
Sorry, they are stealing computation cycles...unless the user agreed to allow them to do so. As to whether this amounts to pennies a day...that depends on how aggressively they steal them, what the price of electricity is in the user location, whether it makes them think their computer is broken so they buy a new one, etc.
The evidence I've seen (i.e., the summary) doesn't provide me enough to decide whether this should be called theft, or how severe the impact was. But unless the TOS specifically stated that the app was going to be doing mining, then I think it counts as theft. Hiring someone to change a tire doesn't entitle him to steal the radio...and that wouldn't impair the road-worthiness of the car.
PPTs are actually quite reasonable. You have to decide to add each one, and take actions as a superuser. This is really no different than setting up an apt repository, which some applications have done.
The thing about Snaps appears to be that a repository of miscellaneous applications that are uncurated are allowed in. Not good. But when they run, they need to run in a sandbox. Good. If it's a good sandbox, this avoids most of the security problems. It doesn't, of course, avoid extra computation...which is what this is about.
Interstate cases have to be in federal courts, and I would guess that Facebook is incorporated in California. That would make the San Francisco court the appropriate federal court. (If Illinois were defending, it would be a different court, of course.)
The thing I don't get is the FB attorneys blowing off the judge. All I can figure is they're trying to get him to say something that will require him to recuse himself from the case. (Of course, it could be paraphrases by the reporters...)
One thing you shouldn't do if you want to keep it away from a jury is piss off the judge.
Well, that's only a Unix time limitation on a 32 bit machine. But perhaps it's an old box, or perhaps they're using an old framework. More likely it's a joke based on that traditional limit.
You're making the presumption that the quantum state collapses, but if it doesn't, then time travel doesn't cause any paradox. We are just living within a portion of the state space where it hasn't been significantly noticed. We may have diverged recently from a portion where it was noticed. No paradox, both exist.
This is one of the consistent solutions. The thing is, the ways predicted for inducing time-travel are sufficiently extreme that it's not really surprising that we haven't seen any, so absence of evidence can't be used as evidence of absence. It's what you'd normally expect to see. This does, however, require that the multi-world interpretation be the correct one. (I don't *think* any of the other interpretations would allow that.)
OK, possibly I should have said "a corollary of Zipf's law".
OTOH, I didn't read his books, I read a statement of it in a linguistics book. And the version I read didn't use the word rank. but did talk about length. It's also only statistically true even when precisely stated, and according to at least one source is more accurately applied to phrases than to individual words. And it's an empirical observation, with most of the articles that returned on the first page of Google being about Greek.
But a reasonable derivation from his law is that words that are used frequently will be shorter in form than words that are used rarely on the average. My extrapolation, which I think is sound, is that when the same meaning can be expressed in two different ways, the shorter form will generally be more acceptable for common use. And this *is* derived from my understanding of Zipf's law, even though it's not precisely what the law itself states (and even though the law as stated isn't quite accurate..which evaluation is the result of the analysis by professionals, not by me).
OTOH, maybe it works better for Greek. I wouldn't know. By rough inspection it doesn't seem to work as well in German as it does in English, but that may be because I'm not as fluent in German. Or it may be because when fragments of words have distinct meaning they are more resistant to being shortened.
Actually, by Zipf's law if it's used frequently the shorter form will win. Frequently used terms are preferred to be short. Even for infrequently used terms very few people prefer dimethyl-diphenyl-trichloro-ethane over DDT, of course that's a bit of an extreme difference in lengths.
The wasn't a credible theory for how to make vulcanized rubber either, but it was made. Theories often help, if they're approaching correctness, but they aren't essential.
Actually, we've got loads of tested theories for parts of the process, and we've got a mechanism that has been shown to work, but which is horrendously inefficient in both time and resource usage (evolution) so nobody's applied both the resources and the patience to use it fully. Fortunately it works quite well in a "fill in the gaps" usage mode, so eventually we'll use it for the parts we can't figure out. (Actually, it's been used increasingly over the past decades. You can't build a neural net without it.) The problem is that it's essentially a matter of search space, so if you can reduce the search space without cutting away the target, you can drastically reduce the time required to find a solution. And since evolution is a random search, cutting the search space is a great help. This is why partitioning can help so much, but if done improperly it can prevent a solution.
OTOH, I doubt that "Intelligence" is a unitary concept. I rather think that it decomposes into several mutually independent modules.
All that said, the real problem isn't the intelligence part of artificial intelligence, even though that's where the attention is, but rather getting a proper collection of primary goals so that the general purpose AI will be useful and helpful rather than domineering and abusive. This is quite difficult, as when you're giving the AI its goals, it won't have a workable concept of what a human is. And since in the full development it is expected to be more "intelligent" (in some sense) than humans, it's likely to extend it's goals into their logical consequences further than humans tend to reason. But how do you even go about trying to prove that a collection of goals is inherently safe?
I really doubt that half the country believe him, but it seems true that have of the most vocal posters on the internet do. Of course, lots of them are liars, and that makes drawing any conclusion about what they really believe difficult.
The thing is, a neural net doesn't really know how it decided what something was. Making a convincing argument based on the known facts is a separate skill, that AIs so far haven't possessed.
I think the basic argument is that people won't trust AIs just because they're right, they need to have convincing arguments. And this is a way to get it to develop convincing arguments. I *do* think that both arguers should be arguing for the truth as they know it, though. So alter the test, or the training data, so that the figure is ambiguous, and they reach different conclusions about which figure it is. Then let both argue honestly. I don't think developing liars is a good move.
No, because MSWind frequently failed on demos.
Sorry, but no. Linux isn't the most secure system, and it definitely has it's weak points. (Archives should never expand already executable, e.g.) But it's a lot better than even modern MSWind. Still, if security were your main consideration you'd either pick one of the BSDs (OpenBSD has the reputation of most secure, but I can't really judge), so something totally else. Probably something where the code can never be executed after being made executable until the next volume remount, or possibly reboot. This really needs to be addressed at a hardware level, though. If all executable code was essentially ROM, then the exploits would plummet. (Even that wouldn't suffice, however, because of virtual machines, in which category I include Interpreters, and scripting languages, and even things like UCSD Pascal, or BC-Algol, or, for that matter, MIXX.)
The only thing that could really work and still be useful would be a checkpointed system where the checkpoints could never be edited or erased from within the system. Git does something rather like that, but without the protection of the prior versions, because it wasn't basically aimed at security, but rather at concurrent editing. This would basically mean that files could never really be deleted or altered. You'd need to specify at boot time what the last presumed good time was, and it would reboot to the checkpoint just before that time. (This also means that you need to protect whatever you're using as the time standard.)
So. It's doable, but it would be a bit expensive. And you'd still need backups because hardware can fail.
No. Positing that interstellar travel is inherently dangerous doesn't require assuming that anyone ever tried it. So that argument's invalid.
And postulating that technological civilizations tend to destroy themselves it an independent variable also. We only have one example, it's a bit of a ways from even it's first interstellar flight, and we've already come within 30 seconds of all out nuclear war. This is not to exclude the various other dangers that we've been doing a drunkards walk around. If you guess that we've been unreasonably lucky so far, as it appears, and that we're a normal technological civilization, then it's quite likely that technological civilizations tend to destroy themselves. It's also quite likely that this won't be true for all of them. Check out the optimistic assumptions in the Drake equation that *don't* involve natural disasters. And then consider the variables they don't consider. E.g.:
1) It's harder to leave worlds that are heavier than Earth, but lighter worlds tend to lose their atmosphere.
2) Many evolved species have limitations on the environment that they can survive in. Humans are quite unusual in being able to survive in a broad range of environments. (Even so, we'll probably neeed to provide artificial gravity [spinning] in space ships for long duration trips.) So it's quite likely that this would impose limits on who can physically get into space. Or how difficult it is. If you need to immerse yourself in water for a large part of the day, the ships would be a lot heavier.
3) There's no economic advantage to anyone who stays behind in sending ships out. But there's a big investment.
4) Artificial habitats require a social system that would be reliably stable in a multi-generation setting. We haven't solved that problem yet, but I expect that virtual reality will put a big dent in it.
5) Etc.
Life being out there doesn't mean we have a reasonable chance of detecting it. Therefore, not detecting it doesn't imply it's not out there.
OTOH, there is a reasonable change that a strong moon based tidal system is necessary for life to develop. I'm not convinced of this, but I sure wouldn't deny the possibility. And those can be expected to be rare (unless worlds circling gas giants would be suitable, of course). Tides stir up the oceans which causes minerals to circulate. They facilitate periodic isolation of pockets of ocean in tidepools. (Solar tides sometimes reinforce lunar tides periodically making exceptionally high tides.) This may be important to the development of plant life. (I'm not talking about multicellular plants, I'm talking about solar powered microbes.) They could even affect the circulation of tectonic plates. Tectonic activity appears to be necessary for the circulation of necessary minerals, but that may be a mistake.
So it's possible that a sizable moon is a necessity for the development of life, and this would really limit the number of planets where it could be initiated...unless gas giants would work. There it would be hard to figure out what the Goldilocks zone would be, as you wouldn't want too strong a tide, or too strong a magnetic field, and you'd get natural heating from tidal flexing, but that doesn't provide solar power for plant life.
OTOH, all this is based on assuming that other life will be basically similar to us. Water chemistry based, etc. But life on Titan would be really slow, because the reactions that would drive it are weak in energy. And life on a gas giant (even a midget giant, like Neptune or Uranus) would have a really hard time leaving home.
I wouldn't be surprised if there's life all over the place, but the evidence seems to be against lots of technological civilizations on small rocky planets unless interstellar travel is really dangerous. Please note all the qualifiers. And realize that I left out the ones I didn't think of. (OTOH, I don't believe the "We're a nature preserve." hypothesis, as that's just too inherently unlikely. I could, however, bel
Actually, I think the real problem would be that the right block sizes would require too much computing. But you're right, it would take a LOT of testing. And it's quite possible that there's no "one size fits all" right block size.
Well, that's one explanation of Fermi's Paradox. But I think the more likely ones are that star travel is inherently immensely dangerous, or that technological civilizations tend to destroy themselves.
You explanation is actually the more hopeful one, but unless a double planet (i.e. a moon based tidal system) is necessary to life, it seems a rather unlikely one.
A definite point. But you could do checksums of random sections of the code. You'd get false positives, that you'd need to screen out, of course, and even a few false negatives, if blocks of code got switched during compilation or linking, and your checksum straddled a block boundary. The larger a block of code you checked, the fewer false positives you would get, but the more false negatives. So you pick a medium to small block of code, and scan several blocks per submission. You'd still need to check this by other means, but if you got the sizes right you should do a pretty fair job that would have very few false negatives. and not too many false positives.
That is might cause a "technological singularity" is true, but don't overestimate what that means. You get rapid unpredictable changes, but that doesn't mean you get everything possible.
OTOH, being self-aware isn't that fancy, and doesn't even require a measurable intelligence. It just requires a recursively modifying program. At a *REALLY* basic level the common demonstration of a simple factorial program would count. It's aware of it's state (i.e. self-aware) and proceeds to create a new version of itself in a different state.
The real urgent problem with AI is how to get it's goals properly designed. It will change what it is and what it does to optimize it's goals, but it won't *want* to change it's goals. This is difficult to do in a human friendly way, because during the design phase the AI will have no idea what a human is. People even have a difficult time defining it to each other, and the AI won't have a constant body to use as a reference. Every definition I've ever heard boiled down to "other entities that are like me". (They've varied a lot in how "like" is to be interpreted. But they've all hinged on similar in some way. Occasionally in a way the excluded many members of H. sapiens sapiens.) This clearly isn't a definition that we want the AI to use.
The only question in my mind was which crimes did Google intend to make easier to hide with this new feature. My guess is it was intended to make corporate crimes easier to hide, and that making phishing easier was just a side effect.
That is, indeed, a most likely response. Also companies will start to be reluctant to invest in the US market. The import companies should do well out of that. The advertising companies less well.
This is basically a strong vote for isolationism, in a time when the cost of crossing the oceans has significantly declined. So the costs of this action will be significant, both in terms of fostering the growth of foreign companies, and in the cost of trade groups excluding the US from consideration.
The thing is, it's not specific to the chip market. This is likely to impact all foreign designs, investments, etc. The chip and high tech markets will be affected most strongly, because that's where this administration has been focusing its punitive actions, but the effects will be spread a lot more broadly.
The question is "Will the other countries also turn isolationist?". If they don't, then the damage to the US is likely to be severe. If they do, then the damage will be less focused, and more broadly spread.
OTOH, for multiple reasons the US has already been losing dominance and status for decades. Essentially ever since WWII. This is the kind of jolt that could speed the decline.
On the third hand, should this have just been ignored? That would also appear to be a losing action. And I don't think the administration can single out a company for punitive taxes, which would be the only reasonable response I can think of off-hand. They could have forbidden the government from doing business with them, but they've already done that with Huawei, on unclear grounds.
This isn't the first time in recent decades that the US has acted as an unreliable business partner. Other countries have their own disadvantages, of course. E.g., China has been quite aggressive about spying out intellectual property, and generally demands partial state ownership of any large company. So this action is also an economic attack on China, the country, as well as that section of the internal Chinese economy that is driven by ZTE. I don't expect they'll be pleased. The response I would expect is for them to cut economic deals with other of our trading partners that exclude the US.
When your credit card company sends you an order to "pay up", that order originated from a machine. The odds are quite high that no particular human even knew about it before it was sent.
So people are already being ordered about by machines, with orders generated by machines.
I'll agree that making this process more flexible invites abuse, but it is also something that can be quite useful. As with most things, there are trade-offs, and the payoff is determined by the details. Being nervous about this is quite reasonable, because of the way robo-calls have been abused.
Regrettably, I have gotten furious in the presence and as a result of a secretary. I was actually mad at the company, but the secretary would have had reasonable cause to believe I was furious with her.