Actually, for even a lousy version of what they're promising the computational costs would be high enough that a local version would need a high end server to run it.
I expect that what they're really planning on is "cloud-sourcing" the answers and speech recognition. And adding features as time goes on, so what they're promising are things that have even yet been designed.
So, yes, this *does* need to be "cloud-connected" to work in an even marginal way. This doesn't mean your ancillary projections are wrong, though. Or that they won't become correct if the device proves successful. (And if it isn't it's likely to suddenly stop working without warning.)
It's not just a table, it comes with an attached keyboard that (appears to be) full size. But it doesn't seem to have usb ports. So I'll probably give it a miss, just like I have the other chromebooks.
I'm rather strongly interested in a really portable computer...but this doesn't seem to be it. I'm just judging by their ad, of course.
Someone else called it a dumb terminal. I think that's a bit overly harsh, but it doesn't seem to be a real computer, either.
An earlier story claimed that the original hacking group passed it on to a different set of hackers when they figured out how valuable it was. I interpreted that as meaning they'd already sold the access, the story figured that was proof it was a nation-state.
So you can take your pick. Either it's already on the market, or it's in the hands of an unknown nation state...from which it will likely leak (eventually), because it won't hurt their employers, and somebody always needs more money.
The problem is, when someone, say Equifax, collects the "other evidence useful for proof of identity", then they can impersonate you to anyone who don't personally know you. And if they share that information with some other entity, willingly or not, THAT entity can no impersonate you to anyone who doesn't know you.
We aren't just talking about one piece of information here.
Pancreatic cancer is almost always untreatable by the time it is detected. It's probable that his choice of alternative medicine caused an early death, but likely he chose it because he was told that the standard treatment meant preparing to die, so he went after a long chance. (Admittedly, I don't know his personal beliefs, but I also don't trust the news reports that much. They are processed for entertainment value.)
OTOH, it's quite difficult for someone even similar to him to pick a suitable replacement. I can't at the moment think of any historical example where that has happened. Certainly not at HP. IBM wasn't ever lead by someone of the same stripe, despite how Watson has been idolized in later times.
I used both System 9 and OSX 10.1. OSX was dramatically better. I did keep the System 9 emulation package around because some things, like the tiff encoder, kept needing it. But honestly, I tried to avoid it as much as possible.
That said, there may have been some ways in which System 9 was better than OSX, I'm just not familiar with any of them.
You might be able to come up with reasons why you prefer Apple, but the choice isn't objective. I happen to be perfectly satisfied with Linux. Well...that's not true, I'd prefer gnome2 over the KDE desktop I currently have, but Mate isn't as good....
That's written as if it were an objective choice, but it isn't. I've got my preferred way of doing things, and that's my preference. Nothing says you should have the same preferences. I do happen to think that gnome3 is objectively worse than gnome2, but there are those that disagree with me. The choice of Mate vs KDE vs gnome2, however, is driven by personal choices. They all have flaws, but the ones in KDE bother *me* less.
Well, unless you believe in actualized infinities any simulation within a simulation would necessarily be simpler. Enough simpler to support the system running the simulation. If so, and if our understanding is approximately correct, we must be near the bottom level of the simulations that can support conscious thought.
I think you're assuming serial computation rather than parallel processing of independent streams that can interact with each other.
FWIW, I suspect that in addition the the problems posed by separate computational units, quantum processing has numerous aspects that we haven't yet noticed.
But even with just parallel streams of computation you can have multiple while(true) loops. it's even been suggested that certain variables have a undefined but bounded value until they are actually accessed, at which time their value become fixed to some (random?) value within those bounds. But if it were pseudo-random rather than actually random how would we know?
There's no plausibility to the Con studies. We can't know the mechanism that would be used to produce the simulation, or what it's inherent features would be. All we can do is list certain minimum capabilities.
OTOH, it's plausible that the Pro studies could come up with reasonable evidence. Not with proof, but with a plausible argument. I do agree that a definite proof is probably impossible.
No. The biggest one probably has ants, cockroaches, or termites at the top. I *think* the bacteria are too fragmented to have a competitor. And I'm judging the size of the food chain by the mass of it's components.
I think the problem is a net of systems. Think of it as a problem in concurrent programming. If you don't know the state of the system, you can't really fix it. Often systems known to be infected need to be recovered from backup. This system sounds as if they didn't *have* any reliable backups.
You are right that it's a matter of culture/priorities, which is why I said things like "you need strong management support". But for anything to work you need to START from a known state.
My first thought was that such tests should be done on an island....but then an early post said that Florida was already testing GM mosquitoes. If that's correct, then, unless this is a different modification, I guess they might as well go ahead.
One proposal I heard was to modify mosquitoes so that they were immune to malaria, but I doubt that's far enough along to be practical.
That isn't even what they said. What they said was "When the hack turned out to be unexpectedly valuable, they turned it over to a more skilled group", and that *this* indicated it was a nation-state.
To me that sounds like they found something valuable and sold it to someone else...who might have been a nation-state. Why not, they have deeper pockets than most.
That may be so, but if you don't start with the system in a known state, any further efforts are possibly worthless. You can't be sure that they're worthless, but you can't be sure that they aren't.
Back in the old days the solution was to set up a duplicate system, update it, test it, and then switch to it. That doesn't work for a highly interactive system. So all that's left as possible is to pick a decent time, say 11PM EST on a Friday, and take the entire system down for "scheduled maintenance". Back it up. Fix everything you know is wrong, test it, back it up again, and then bring it back up. Then, with a system starting from a known state that's as good as can be managed, then you address the more permanent problems. But you need management solidly behind you to do that kind of thing.
I don't know whether they've fixed the problem, as I haven't used it in decades, but I remember it as "The database that couldn't add two number correctly". It had a bunch of other problems, but it was so convenient that I used it until I actually caught it adding two numbers together and getting the wrong answer (repeatedly, but not on every run). After that I transitioned away from it as quickly as I could. It took a lot of testing, as I couldn't believe a business database could be that bad.
It's not *just* that they aren't adequately staffed and funded, though that is also true. Regulatory capture is an even worse problem. No regulator should be allowed to accept any remuneration from those they regulate, not even after they retire from the body. And I mean not allowed to accept *ANY* remuneration. No jobs. No dinners. No speaking fees. No consultant arrangements. No discounted apartments. No payments for stock owned. NOTHING. And not just while regulating, but also afterwards. (If they own stocks or bonds then they better sell it before they take the job as regulator, because afterwards they are allowed neither to collect dividends nor to sell it.)
Being a regulator should entail a final and permanent severance of all connections with those regulated. If the regulatee is the spouse of the regulated, this should not only require a divorce, it should require that they never henceforth exchange any communications. Not even through intermediates...such as children. So in that case don't take the job.
It's totally off-topic, but it does have some merit. The 17 amendment is one of the steps that limited the power of the states and increased the centralized power of the Federal Government. It's plausible that it was a mistake, though it was intended to address real existing problems. And the problems that it caused is one of the things making me hesitant to support efforts to remove the electoral college. I can see the clear problems that it causes, but what I can't see is the problems it prevents.
The problem with purchasing a house without debt is that rent for living space increases in cost without bounds otherwise. And when purchasing a house becomes too difficult, even with debt, the cost of rental housing explodes, even though there is a practical bound. As it has recently in many cities of the US, and probably elsewhere.
This doesn't say what the answer to that problem should be. It's merely constraints on the answer. Certainly rent-control has many well known problems, and has proven an undesirable answer...though possibly not as undesirable as arbitrarily increasing rents.
It's my suspicion that the basic problem can be traced to an exploding population combined with housing that degrades in quality quickly. This combination makes any real solution impossible, as land area within easy reach of an urban area is limited. Faster transport tends to lead to larger cites, and a denser urban core, with higher prices. Rapidly degrading housing quality tends to cause that urban core to degrade in quality, but the demand keeps prices from decreasing as the quality degrades.
Once upon a time the cities were where the rich lived, and the poor lived in the suburbs. This is because slow transport meant that it took a long time to get anywhere. I suspect that the time I was reading of was shortly after a war had destroyed much of the city and it had been freshly rebuilt. But the pressure of "commute time" remains. So you really need to factor the time spent commuting into the cost of the housing.
The only answers I've run across that were at all convincing were fictional. The idea of "archology" with elevators and escalators for rapid 3-dimensional transport has it's plausibility. But the cost wouldn't be trivial, and currently the maintenance would probably be impossible. And maintenance would be needed, or the entire thing would become unlivable. But note that the only plausible archology stories that I've read also feature an extreme degree of surveillance by the organization that runs the archology. Sometimes it's benevolent, but this is a bit dubious because people with unrestrained power don't tend to be benevolent. (Some have said it's not power that corrupts, but the lack of consequences for it's misuse. This is true, but if you have power one of the first things you will do is protect yourself against consequences for inadvertent mistakes...and that's assuming you start from an ethical and benevolent position.)
It hasn't been proven that the particular officials that engaged in stock trades had, or even should have had, knowledge of the incident. That I believe they did is irrelevant.
OTOH, WTF!!! Why on earth would I believe any report made by a company investigating itself for wrong-doing. Even when there's an outside investigation companies are *known* for passing the blame to innocent parties. Or subordinates who were operating under threat, and were at most accessories after the fact.
To me this screams that the board itself was probably participating in this scam. Whether than can be proven or not is a separate matter. I also feel they will definitely hide any evidence of their own guilt as thoroughly as they can.
Yah. And the key point is that even then, when people were more trustworthy (in general), he didn't have access to the inside of the house.
P.S.: The reason that people were more trustworthy is that people grew up in small towns where everyone knew everyone else, and transportation was slow. In fact SLOW. So if something bad happened the pool of suspects was small. (You may recall that people who grew up in such environments traditionally distrusted strangers. This was reasonable. Some people lived in cities, but around 90% of the population didn't. And strangers were suspected of being people who had been ostracized by the people who really knew them.)
No, you're preventing them from being conscious. So imprisonment doesn't fit. It's more like stealing their soul than anything else I can think of, but even Stormbringer didn't ever return the souls to active status. It's the possibility of reincarnation in the same body that makes this so difficult to describe in normal concepts. Come to that, you could also clone the AI's soul into multiple bodies...which aren't all identical.
We don't have the right language to talk about it because nobody has experience with it, and there haven't been enough stories featuring this theme to develop a way of talking about it. (We can talk about FTL starships, even though we don't have any, because there've been lots of stories that developed ways of talking about them. But go back to the 1930's [e.g. E.E.Smith's "Skylark of Space" and you'll notice that he has a hard time talking about it.)
Slashdot Mirror
The thing is, it's probably designed to update itself over the net.
So, the first version may well be secure...and minimally capable. One of the updates, though, will open the floodgates.
Actually, for even a lousy version of what they're promising the computational costs would be high enough that a local version would need a high end server to run it.
I expect that what they're really planning on is "cloud-sourcing" the answers and speech recognition. And adding features as time goes on, so what they're promising are things that have even yet been designed.
So, yes, this *does* need to be "cloud-connected" to work in an even marginal way. This doesn't mean your ancillary projections are wrong, though. Or that they won't become correct if the device proves successful. (And if it isn't it's likely to suddenly stop working without warning.)
It's not just a table, it comes with an attached keyboard that (appears to be) full size. But it doesn't seem to have usb ports. So I'll probably give it a miss, just like I have the other chromebooks.
I'm rather strongly interested in a really portable computer...but this doesn't seem to be it. I'm just judging by their ad, of course.
Someone else called it a dumb terminal. I think that's a bit overly harsh, but it doesn't seem to be a real computer, either.
An earlier story claimed that the original hacking group passed it on to a different set of hackers when they figured out how valuable it was. I interpreted that as meaning they'd already sold the access, the story figured that was proof it was a nation-state.
So you can take your pick. Either it's already on the market, or it's in the hands of an unknown nation state...from which it will likely leak (eventually), because it won't hurt their employers, and somebody always needs more money.
Are you claiming that Equifax is better? Or just that Equifax doesn't make their mistakes public?
The problem is, when someone, say Equifax, collects the "other evidence useful for proof of identity", then they can impersonate you to anyone who don't personally know you. And if they share that information with some other entity, willingly or not, THAT entity can no impersonate you to anyone who doesn't know you.
We aren't just talking about one piece of information here.
Pancreatic cancer is almost always untreatable by the time it is detected. It's probable that his choice of alternative medicine caused an early death, but likely he chose it because he was told that the standard treatment meant preparing to die, so he went after a long chance. (Admittedly, I don't know his personal beliefs, but I also don't trust the news reports that much. They are processed for entertainment value.)
OTOH, it's quite difficult for someone even similar to him to pick a suitable replacement. I can't at the moment think of any historical example where that has happened. Certainly not at HP. IBM wasn't ever lead by someone of the same stripe, despite how Watson has been idolized in later times.
I used both System 9 and OSX 10.1. OSX was dramatically better. I did keep the System 9 emulation package around because some things, like the tiff encoder, kept needing it. But honestly, I tried to avoid it as much as possible.
That said, there may have been some ways in which System 9 was better than OSX, I'm just not familiar with any of them.
You might be able to come up with reasons why you prefer Apple, but the choice isn't objective. I happen to be perfectly satisfied with Linux. Well...that's not true, I'd prefer gnome2 over the KDE desktop I currently have, but Mate isn't as good....
That's written as if it were an objective choice, but it isn't. I've got my preferred way of doing things, and that's my preference. Nothing says you should have the same preferences. I do happen to think that gnome3 is objectively worse than gnome2, but there are those that disagree with me. The choice of Mate vs KDE vs gnome2, however, is driven by personal choices. They all have flaws, but the ones in KDE bother *me* less.
Well, unless you believe in actualized infinities any simulation within a simulation would necessarily be simpler. Enough simpler to support the system running the simulation. If so, and if our understanding is approximately correct, we must be near the bottom level of the simulations that can support conscious thought.
I think you're assuming serial computation rather than parallel processing of independent streams that can interact with each other.
FWIW, I suspect that in addition the the problems posed by separate computational units, quantum processing has numerous aspects that we haven't yet noticed.
But even with just parallel streams of computation you can have multiple while(true) loops. it's even been suggested that certain variables have a undefined but bounded value until they are actually accessed, at which time their value become fixed to some (random?) value within those bounds. But if it were pseudo-random rather than actually random how would we know?
There's no plausibility to the Con studies. We can't know the mechanism that would be used to produce the simulation, or what it's inherent features would be. All we can do is list certain minimum capabilities.
OTOH, it's plausible that the Pro studies could come up with reasonable evidence. Not with proof, but with a plausible argument. I do agree that a definite proof is probably impossible.
No. The biggest one probably has ants, cockroaches, or termites at the top. I *think* the bacteria are too fragmented to have a competitor. And I'm judging the size of the food chain by the mass of it's components.
I think the problem is a net of systems. Think of it as a problem in concurrent programming. If you don't know the state of the system, you can't really fix it. Often systems known to be infected need to be recovered from backup. This system sounds as if they didn't *have* any reliable backups.
You are right that it's a matter of culture/priorities, which is why I said things like "you need strong management support". But for anything to work you need to START from a known state.
My first thought was that such tests should be done on an island....but then an early post said that Florida was already testing GM mosquitoes. If that's correct, then, unless this is a different modification, I guess they might as well go ahead.
One proposal I heard was to modify mosquitoes so that they were immune to malaria, but I doubt that's far enough along to be practical.
That isn't even what they said. What they said was "When the hack turned out to be unexpectedly valuable, they turned it over to a more skilled group", and that *this* indicated it was a nation-state.
To me that sounds like they found something valuable and sold it to someone else...who might have been a nation-state. Why not, they have deeper pockets than most.
That may be so, but if you don't start with the system in a known state, any further efforts are possibly worthless. You can't be sure that they're worthless, but you can't be sure that they aren't.
Back in the old days the solution was to set up a duplicate system, update it, test it, and then switch to it. That doesn't work for a highly interactive system. So all that's left as possible is to pick a decent time, say 11PM EST on a Friday, and take the entire system down for "scheduled maintenance". Back it up. Fix everything you know is wrong, test it, back it up again, and then bring it back up. Then, with a system starting from a known state that's as good as can be managed, then you address the more permanent problems. But you need management solidly behind you to do that kind of thing.
Access? Are you talking about MSAccess?
I don't know whether they've fixed the problem, as I haven't used it in decades, but I remember it as "The database that couldn't add two number correctly". It had a bunch of other problems, but it was so convenient that I used it until I actually caught it adding two numbers together and getting the wrong answer (repeatedly, but not on every run). After that I transitioned away from it as quickly as I could. It took a lot of testing, as I couldn't believe a business database could be that bad.
It's not *just* that they aren't adequately staffed and funded, though that is also true. Regulatory capture is an even worse problem. No regulator should be allowed to accept any remuneration from those they regulate, not even after they retire from the body. And I mean not allowed to accept *ANY* remuneration. No jobs. No dinners. No speaking fees. No consultant arrangements. No discounted apartments. No payments for stock owned. NOTHING. And not just while regulating, but also afterwards. (If they own stocks or bonds then they better sell it before they take the job as regulator, because afterwards they are allowed neither to collect dividends nor to sell it.)
Being a regulator should entail a final and permanent severance of all connections with those regulated. If the regulatee is the spouse of the regulated, this should not only require a divorce, it should require that they never henceforth exchange any communications. Not even through intermediates...such as children. So in that case don't take the job.
It's totally off-topic, but it does have some merit. The 17 amendment is one of the steps that limited the power of the states and increased the centralized power of the Federal Government. It's plausible that it was a mistake, though it was intended to address real existing problems. And the problems that it caused is one of the things making me hesitant to support efforts to remove the electoral college. I can see the clear problems that it causes, but what I can't see is the problems it prevents.
The problem with purchasing a house without debt is that rent for living space increases in cost without bounds otherwise. And when purchasing a house becomes too difficult, even with debt, the cost of rental housing explodes, even though there is a practical bound. As it has recently in many cities of the US, and probably elsewhere.
This doesn't say what the answer to that problem should be. It's merely constraints on the answer. Certainly rent-control has many well known problems, and has proven an undesirable answer...though possibly not as undesirable as arbitrarily increasing rents.
It's my suspicion that the basic problem can be traced to an exploding population combined with housing that degrades in quality quickly. This combination makes any real solution impossible, as land area within easy reach of an urban area is limited. Faster transport tends to lead to larger cites, and a denser urban core, with higher prices. Rapidly degrading housing quality tends to cause that urban core to degrade in quality, but the demand keeps prices from decreasing as the quality degrades.
Once upon a time the cities were where the rich lived, and the poor lived in the suburbs. This is because slow transport meant that it took a long time to get anywhere. I suspect that the time I was reading of was shortly after a war had destroyed much of the city and it had been freshly rebuilt. But the pressure of "commute time" remains. So you really need to factor the time spent commuting into the cost of the housing.
The only answers I've run across that were at all convincing were fictional. The idea of "archology" with elevators and escalators for rapid 3-dimensional transport has it's plausibility. But the cost wouldn't be trivial, and currently the maintenance would probably be impossible. And maintenance would be needed, or the entire thing would become unlivable. But note that the only plausible archology stories that I've read also feature an extreme degree of surveillance by the organization that runs the archology. Sometimes it's benevolent, but this is a bit dubious because people with unrestrained power don't tend to be benevolent. (Some have said it's not power that corrupts, but the lack of consequences for it's misuse. This is true, but if you have power one of the first things you will do is protect yourself against consequences for inadvertent mistakes...and that's assuming you start from an ethical and benevolent position.)
Not true!! I believe in the flying spaghetti monster, lord of the universe!
It hasn't been proven that the particular officials that engaged in stock trades had, or even should have had, knowledge of the incident. That I believe they did is irrelevant.
OTOH, WTF!!! Why on earth would I believe any report made by a company investigating itself for wrong-doing. Even when there's an outside investigation companies are *known* for passing the blame to innocent parties. Or subordinates who were operating under threat, and were at most accessories after the fact.
To me this screams that the board itself was probably participating in this scam. Whether than can be proven or not is a separate matter. I also feel they will definitely hide any evidence of their own guilt as thoroughly as they can.
Yah. And the key point is that even then, when people were more trustworthy (in general), he didn't have access to the inside of the house.
P.S.: The reason that people were more trustworthy is that people grew up in small towns where everyone knew everyone else, and transportation was slow. In fact SLOW. So if something bad happened the pool of suspects was small. (You may recall that people who grew up in such environments traditionally distrusted strangers. This was reasonable. Some people lived in cities, but around 90% of the population didn't. And strangers were suspected of being people who had been ostracized by the people who really knew them.)
No, you're preventing them from being conscious. So imprisonment doesn't fit. It's more like stealing their soul than anything else I can think of, but even Stormbringer didn't ever return the souls to active status. It's the possibility of reincarnation in the same body that makes this so difficult to describe in normal concepts. Come to that, you could also clone the AI's soul into multiple bodies...which aren't all identical.
We don't have the right language to talk about it because nobody has experience with it, and there haven't been enough stories featuring this theme to develop a way of talking about it. (We can talk about FTL starships, even though we don't have any, because there've been lots of stories that developed ways of talking about them. But go back to the 1930's [e.g. E.E.Smith's "Skylark of Space" and you'll notice that he has a hard time talking about it.)