I'm not familiar enough with the license they chose, but does it guarantee patent protection? The thrust that MS is currently using against FOSS seems to depend on software patents. If they had chosen the GPL, or GPL3, or BSD, or AGPL I would have an idea of what the significance was, but Creative Commons isn't commonly used for FOSS software, so I don't know what that means as far as patents. (WRT copyrights I can make fair guesses, but that's a different matter.)
On a long enough time span, you are probably correct. OTOH, the ecology of the Gulf had already been stretched near the breaking point, the dead zones were already growing, etc.
I'm not sanguine as to the outcome of this latest problem. It may only lead to a few thousand deaths, and a few hundred thousand who are sick (perhaps mildly) for the rest of their lives. But that's just people. The fisheries may never recover. (They were already pretty stressed, and fisheries that collapse often don't recover for at least centuries, and perhaps never.) When a major species of fish dies, something else moves in, often something that people don't find as appealing. (We are, after all, one of the major stresses on a species.)
I don't think that one can judge just how serious this event is by looking only at a temporal snapshot. And things don't usually have only one cause. The proverbial "straw that broke the camel's back" is an example, but this is more like a sedan chair. Not enough to cause a major disaster by itself, but not a minor factor, either.
You're fooling around with methyl hydrates. Yes it could be worse than BP. The Gulf disaster was relatively localized. Methyl hydrates have a reputation for large areas spontaneously releasing, even without intervention. (There's reasonable grounds for suspicion that some tsunami have been caused by such spontaneous detonations.)
Well, if it's done carefully and safely, afterwards things might be better. (There's a notable chance, after all, that if the sea warms much the rate of spontaneous detonations will increase.) But the problems (that occur to me) are: 1: Can you avoid touching it off while trying to extract it, and 2: What happens to the carbon after the methane is burned. (Admittedly, that's no worse than coal, and it wouldn't have the problem with soot that coal does.)
Previously US companies have looked at methyl hydrates, and the problems involved, and backed away.
Yes, there are patent trolls, and the system does facilitate them. It doesn't, however, facilitate a real small competitor. Patent trolls don't dare do anything BUT troll, because if they did they would become vulnerable to counter-suits for patent violation.
*This* is a good system?
Well, yes, if you're a major corporation. Patent trolls are a major headache, but you can often buy them off, and send them chasing your competition. They're good for major corporations in the same way that cows are good for grass: They kill off the competition.
Sorry, but many people *did* notice the *break the net* aspect of the JavaScript mods immediately. I'll grant you that many didn't, and, in fact, many still don't. This doesn't mean it wasn't done with malice aforethought, or that the vile results weren't predictable from the start.
That's not quite the way I remember things happening.
OS/2 vs. Windows was MS modifying a work-for-hire and then claiming it as their own, while IBM was bound by a consent decree, and so didn't legally protest. (They *were* more interested in the mainframe business, of course, but they acquiesced rather than giving MS the go ahead.)
Mosaic: IE was "purchased" from Spyglass...for a percentage of the sales profits. Then MS gave it away, so Spyglass died.
Java: MS developed an incompatible dialect of Java called J++ and refused to support Java, despite a contract to do so. (Sun sued them and eventually won.)
P.S.: Wasn't it Netscape that was developed by the people from Mosaic? *Not* IE. (Though to be fair, the Spyglass people may also have been from the Mosaic project.)
The difference is that Java was put under the GPL by the copyright owner. Mono *may* have been. (When Mono first came out an MS executive declared "That software contains our intellectual property, and we will defend it". I have always understood this as an assertion that Mono didn't have clean title to the code that they licensed, and that therefore you cannot rely on the license.)
That suit is tricky, with arguments on both sides. What Oracle is alleging is based around a part of Java that's not GPL. Whether it's valid, I'm not sure, but it's not a clear matter of Good vs. Evil.
On the balance I tend to side with Google in the dispute, but I know quite well that I don't really understand what it's about. IANAL, and only a lawyer would have grounds for claiming that they understood what it was about. But I despise software patents, so my siding with Google doesn't mean much as far as legal status of the case. (I'm pretty sure it's more about patents than copyrights.)
While I don't trust MS, that doesn't automatically make everything they do bad.
My first thought was "Is this legally binding?", closely followed by "Exactly what it covered?". On previous occurrences those were the questions that meant "This is something to avoid.", but that doesn't, necessarily, mean that the same thing is true this time. If it's legally binding, and it is a sufficiently encompassing commitment, then I may give C# a look. But I'll wait until someone I trust analyses the "promise" for both scope and validity...because IANAL, and this looks like something that calls for a lawyer. (E.g.: is it compatible with software licensed under the AGPL? If it isn't, then it's, to me, a useless promise even if it is binding.)
I'd go a bit further, and say you can't have decent language handling without a reasonable model of the world. And that's difficult.
OTOH, emotions are relatively easy. Emotions are basically built-in shortcuts in reasoning or motivation (usually both). As such, the problem is deciding which ones to enable, which to "hard wire", and which to forbid.
E.g. child-fire: The burnt child dreads the fire. Naive thinker encounters attractive object of class O. Naive thinker experiences "pain" on interaction with object. Naive thinker learns "fear interactions with objects of class O" def. fear: Desire strongly to avoid due to expectation of injury.
So now the thinker is less naive, and it has a reasoning short-cut to use WRT objects of class O.
The real problem is that there's LOTS of specialized rules of this sort. Some of them need to be built-in because learning them the hard way would be strongly injurious. There probably aren't many like this, but the ones that are needed are very important. E.g., there was a robot in, I believe Japan, though it might have been Florida, that started disassembling itself. Whoops! And no pain sensors to warn it, either, so there wasn't any obvious way for it to learn in time not to do that.
Note that what rules need to be built-in is strongly dependent on what sensors are available.
Well, according to the latest evidence quantum effects *are* involved in photosynthesis. Not that it couldn't happen anyway, but the efficiency implies that the model using quantum effects to explain the process is the correct one.
Now I don't see any particular reason that this should translate into anything related to thoughts, but I do feel that it means that it's reasonable to consider the matter as undecided until we figure out how things really *do* work. My favorite model of sentience doesn't require, or even suggest, any quantum effects, but there are a few steps where, if quantum effects were available, it could speed things up. (OTOH, I'm not real impressed by the speed of neural processing, so even if there is a possibility of quantum effects, my bet would be that they aren't being used.)
The Turing test is sufficient, but not necessary, for sentience.
Mind you, I'm talking about the formally specified Turing test, not the bunch of half-assed imitations that people throw around as if they were equivalent. (IIRC, the Eliza program once passed a typical informal specification. I.e., it fooled a interrogator into thinking that it was a real person. It also infuriated them to the point where the person who hooked it up to the teletype nearly got fired before he convinced his boss that he had been "talking" to a machine.)
But biological machines (e.g., people) have been optimized for quick replication rather than for durability. Just as an example, the DNA code could be a lot more error-correcting than it is. But it would slow down replication too much, so it's not selected for by evolution.
OTOH, there's no essential difference between a designed biological form and a nano-machined entity. It's possible that designed biological life could be a lot more durable than one derived from mechanical basis. But not necessarily. Easy to evolve and efficient in operation are just about orthogonal axes. Once you start talking about designed biological forms, however, you leave behind the constraints of "easy to evolve", and we don't know what that design space is like. E.g., perhaps a better storage ROM than DNA is available, one that has a lower error rate, and is usable over a wider thermal range. We don't even know how much of current DNA codes are useless. We call a lot of stuff "junk DNA", but that may just mean that we haven't yet figured out what it does.
So I'm not willing to presume a particular distinction between nano-machined life forms and designed biological life forms.
Right or Left, it doesn't make much difference. They both lie freely. Allowing a "news program" to intentionally misrepresent the truth was a very bad decision. Don't know what would have been a good one, though. If you're vulnerable if you present lies, then someone will say that you lied, and probably be able to prove it. (Nobody can be accurate all the time.) I've generally given up on the media as a source of truth, and depend more on bloggers. And THEY lie. But they're much less professional about it, so the lies are usually easier to spot.
It won't just be more of the same, because the environment is changing. The prevalence of automated warehouses changes what's expected of automation. So does the initial appearance of cars that can park themselves, and the prototypes of cars that can drive themselves. Already we are starting work on embedding a model of the external world into the computer. It would be a mistake to all those things intelligent...exactly...but they make information available to a controlling intelligence that wasn't available in prior times. And just think of the competitive advantage an intelligent voice answering system would be over the current menu tree (whether it recognized your voice commands or not). So when someone learns to build one, there will be a ready market. The problem here isn't intelligence. It doesn't take much in the way of intelligence. It's building an abstract model of the external world. Building such a model DOES require intelligence, but updating such a model requires a lot less intelligence, so it only needs to be built twice. At that point a evolutionary "arms-race" will start, and development can be predicted to be rapid.
So one of the basic questions is "How intelligent a model of the world does a system require in order to out-compete a phone system menu-tree?" My assertion is "not much".
For reasons like this I think that the problem is the building of mental models, rather than intelligence per se. (And I'd assert that Bayesian reasoning is pretty close to the definition of intelligence. It's just that intelligence isn't worth all that much by itself. It needs tools to work with.)
It's one of about four or five conventional mechanisms. Generally each of the versions lead into the others, however, so there's no real conflict between them.
E.g., neural interfaces connected via a development of the Internet could yield an "overmind" capable of addressing problems that cannot now be addressed.
Or, those same neural interfaces connected to a computer via an advanced programming interface could enable the development of programs not currently possible by using feedback to stabilize one's thoughts while editing. And this could be used to...
Or molecular biology & genetic engineering could yield more intelligent people who...etc.
Or...
Plug in whatever your favorite technological improvement is, and see how it advances the rate at which things change.
The singularity happens when the rate of change becomes too fast for us to deny that we can't keep up with it. I'd say it's well underway, but I don't expect it to arrive in full blown form until sometime after 2020, and I project 2030 as when it runs totally out of control (i.e., under it's own control... or just uncontrolled, depending on exactly what form it takes).
If we had a government that was ethical and gave serious concern to the quality of the lives of it's citizens, then I'd be doing everything I could to slow down the arrival of the singularity. (Bar forcing civilization to collapse.) As it is, however, if we can manage to live through the transition phase, the singularity may be the only hope of survival for humanity. (My estimate of the odds isn't cheery, hovering around 40%, but my estimate if people stay in control of civilization heads towards zero within a century or two.)
We may not know the limits, but we can put bounds on them. E.g. an upper bound on the number of bits mapped by the brain is the powerset of the number of synapses in the brain (at any one time, but even though the number varies, it doesn't matter which time you pick. It's a pretty generous upper bound.)
I'm sure that someone could justify a much tighter bound, but I don't want to get into quibbles about frequency vs. pulse domain coding of neuron firings, etc., and I'm well outside my specialty.
I didn't read all your post, because I don't agree with your basic concept. I suspect that intelligence is rather simple, though just how simple I'm not certain. And I suspect that there are many approached to intelligence. The neural net is the only one for which we have an existence proof, but that doesn't imply that it's the only possibility.
OTOH, to interface an intelligence usefully with the real world (i.e., to enable it to understand the external reality) is probably much trickier. But that's getting extensive work all the time.
Another problem, and one that probably can't be solved until after the intelligence has a reasonable model of the external world, is language. This is separate from intelligence, but definitely acts as a facilitator for increased intelligence. (It also tends to induce massive blind spots. So there needs to be a secure model of the external world before it is introduced. People tend to introduce language too early in the process, or at least so I believe. This leads to the creation of beliefs in externals for which there is no creditable evidence. Like faeries and other such entities.) Ideally one would introduce epistemology before introducing language, but that's probably impossible. Bayesian reasoning is probably as close as we will be able to come, but it's been proven that given separate sets of priors one cannot assume that additional evidence will cause two reasoners initially different to converge. It's just the best we can do. (But it's better, i.e., more accurate, then human reasoning. OTOH, it's a lot slower. We have evolved to prefer short-cuts that lead to quick decisions over extensive evidence examination.)
So things ARE progressing rapidly towards AI. You many not notice all the changes, or even recognize them when you encounter them, but they are progressing rapidly. I still expect an AI as intelligent as a human by 2030. I'm far less certain about whether this will be desirable. That depends a lot on what it's motivational structure is. And that's something I haven't noticed much progress on...which I consider quite dangerous. (OTOH, lots of the details are being done by companies which are keeping their advanced work secret. You decide whether you trust that dynamic.)
If there's a treatment, then it might be useful to know early if you're going to get it. If there's no treatment, it only means that your health insurance is going to start looking for excuses to cancel your policy.
Students aren't in any position to reform the system, so sabotage, by one means or another, is their only reasonable response. Many teachers seem to feel equally incapable, but they have a few other choices.
(It's unfair to expect every teacher to start a Summerhill, or John Woolman, or other alternative school. But they could change jobs. Given how poorly they are normally treated, I'm surprised that they don't all quit after their first year.)
But for children...just how to you propose that they should fix the problem.
Dynamic braking recovers much of the energy used in accelerating the car.
There are other considerations that make the answer lots more complex, but your particular point isn't a major one. (Mind you, different hybrids make different choices at many points of construction. Some of them clearly AREN'T any better than an efficient gasoline car. For others, it's not clear. Some are clearly more efficient in certain environments, and worse in others. Etc.)
E.g.: The manufacture of the car may well produce more emissions that the running of the car will produce in it's lifetime. So the best "green" statement is likely to be keeping your car running with a well tuned engine for as long as you can. In this context, the batteries of hybrids are pretty much of an unknown. Building them costs and unknown, but large, amount of pollution. How long will they last? Will they be recycled into new batteries? What's the replacement cost? What are the failure modes? (Do they just start not holding a charge?)
For now, I'll continue with a 1990's Toyota Corona. (I think that's the name, I can never keep straight between the Corona and the Corolla. I use whichever is the smaller of the two.) It's only 15 years old so far.
Basically, you're right. In the short term. It needs to become possible to maintain a closed ecology before space colonization is practical, and until then there's no place to go. (Also, don't expect it to be libertarian when it arrives. Things are going to be much too critical for that.)
This doesn't make it less important, but it does mean that expecting it right now is unreasonable. Biosphere 2 rather proved that.
Additionally, it's going to be important to improve the generation of power from solar sources. And energy storage needs to be improved. And extraction of resources from low valued ores...without using water. Etc. Lots of things need to be developed...but the thing is, most of them need to be developed anyway.
When it actually becomes possible, I expect a massive increase in interest. Even before that a permanent Lunar observatory on the far side will probably happen. That will be used to work out the final problems. (But we're aways from when that's practical.)
As to why? I don't know. Political or religious dissidents fleeing repression? Not impossible. Exporting prisoners is unlikely. I think the first colonists will need to be quite wealthy.
A space elevator is probably too ambitious for a first skyhook. Probably we should start with either the star-whale of the pin-wheel. I prefer the pinwheel.)
The pinwheel is an orbital mass with several arms that rotates sufficiently to bring the arms around. It can be in various orbits (and probably should be, to enable one pinwheel to "hand off" to another. For the lowest one, one arm rotates down into the high stratosphere, where a cargo pod is attached. (This may contain people, but it's basically unpowered.) At a certain point, the capsule is released from the arm and is flung off at a target destination. As with all sky-hooks, it's important that as much mass is lowered to the ground as is lifted. (Or it drifts out of orbit.)
The starwhale is an electromagnetic brake (or accelerator). You shoot the payload up to orbital height, but not to orbital velocity. The starwhale swallows it, and the electromagnetic brake accelerates it to orbital speed. It can then act as an accelerator to direct you to any destination you choose. (It may need to be quite long in order to use reasonable G forces for this.) The trick is, you've got to enter the electromagnetic brake properly. This can be quite tricky, especially if you're coming in from, say, Jupiter's moons. You've got a huge velocity, and you need to be properly aligned. If you miss...ugh! Both of you could be destroyed.
As I said, I prefer the pinwheel. There you don't risk destroying everything if there's a slight orbital hangup. (But you'd better have enough delta-V to be able to take another pass at the hook. But that's a much smaller problem.)
Slashdot Mirror
Oh?
I'm not familiar enough with the license they chose, but does it guarantee patent protection? The thrust that MS is currently using against FOSS seems to depend on software patents. If they had chosen the GPL, or GPL3, or BSD, or AGPL I would have an idea of what the significance was, but Creative Commons isn't commonly used for FOSS software, so I don't know what that means as far as patents. (WRT copyrights I can make fair guesses, but that's a different matter.)
On a long enough time span, you are probably correct. OTOH, the ecology of the Gulf had already been stretched near the breaking point, the dead zones were already growing, etc.
I'm not sanguine as to the outcome of this latest problem. It may only lead to a few thousand deaths, and a few hundred thousand who are sick (perhaps mildly) for the rest of their lives. But that's just people. The fisheries may never recover. (They were already pretty stressed, and fisheries that collapse often don't recover for at least centuries, and perhaps never.) When a major species of fish dies, something else moves in, often something that people don't find as appealing. (We are, after all, one of the major stresses on a species.)
I don't think that one can judge just how serious this event is by looking only at a temporal snapshot. And things don't usually have only one cause. The proverbial "straw that broke the camel's back" is an example, but this is more like a sedan chair. Not enough to cause a major disaster by itself, but not a minor factor, either.
You're fooling around with methyl hydrates. Yes it could be worse than BP. The Gulf disaster was relatively localized. Methyl hydrates have a reputation for large areas spontaneously releasing, even without intervention. (There's reasonable grounds for suspicion that some tsunami have been caused by such spontaneous detonations.)
Well, if it's done carefully and safely, afterwards things might be better. (There's a notable chance, after all, that if the sea warms much the rate of spontaneous detonations will increase.) But the problems (that occur to me) are:
1: Can you avoid touching it off while trying to extract it, and
2: What happens to the carbon after the methane is burned. (Admittedly, that's no worse than coal, and it wouldn't have the problem with soot that coal does.)
Previously US companies have looked at methyl hydrates, and the problems involved, and backed away.
Yes, there are patent trolls, and the system does facilitate them. It doesn't, however, facilitate a real small competitor. Patent trolls don't dare do anything BUT troll, because if they did they would become vulnerable to counter-suits for patent violation.
*This* is a good system?
Well, yes, if you're a major corporation. Patent trolls are a major headache, but you can often buy them off, and send them chasing your competition. They're good for major corporations in the same way that cows are good for grass: They kill off the competition.
Sorry, but many people *did* notice the *break the net* aspect of the JavaScript mods immediately. I'll grant you that many didn't, and, in fact, many still don't. This doesn't mean it wasn't done with malice aforethought, or that the vile results weren't predictable from the start.
That's not quite the way I remember things happening.
OS/2 vs. Windows was MS modifying a work-for-hire and then claiming it as their own, while IBM was bound by a consent decree, and so didn't legally protest. (They *were* more interested in the mainframe business, of course, but they acquiesced rather than giving MS the go ahead.)
Mosaic: IE was "purchased" from Spyglass...for a percentage of the sales profits. Then MS gave it away, so Spyglass died.
Java: MS developed an incompatible dialect of Java called J++ and refused to support Java, despite a contract to do so. (Sun sued them and eventually won.)
P.S.: Wasn't it Netscape that was developed by the people from Mosaic? *Not* IE. (Though to be fair, the Spyglass people may also have been from the Mosaic project.)
The difference is that Java was put under the GPL by the copyright owner. Mono *may* have been. (When Mono first came out an MS executive declared "That software contains our intellectual property, and we will defend it". I have always understood this as an assertion that Mono didn't have clean title to the code that they licensed, and that therefore you cannot rely on the license.)
That suit is tricky, with arguments on both sides. What Oracle is alleging is based around a part of Java that's not GPL. Whether it's valid, I'm not sure, but it's not a clear matter of Good vs. Evil.
On the balance I tend to side with Google in the dispute, but I know quite well that I don't really understand what it's about. IANAL, and only a lawyer would have grounds for claiming that they understood what it was about. But I despise software patents, so my siding with Google doesn't mean much as far as legal status of the case. (I'm pretty sure it's more about patents than copyrights.)
While I don't trust MS, that doesn't automatically make everything they do bad.
My first thought was "Is this legally binding?", closely followed by "Exactly what it covered?". On previous occurrences those were the questions that meant "This is something to avoid.", but that doesn't, necessarily, mean that the same thing is true this time. If it's legally binding, and it is a sufficiently encompassing commitment, then I may give C# a look. But I'll wait until someone I trust analyses the "promise" for both scope and validity...because IANAL, and this looks like something that calls for a lawyer. (E.g.: is it compatible with software licensed under the AGPL? If it isn't, then it's, to me, a useless promise even if it is binding.)
Well ... modems used to connect over the serial port. I seem to remember a few viruses that spread that way.
I'd go a bit further, and say you can't have decent language handling without a reasonable model of the world. And that's difficult.
OTOH, emotions are relatively easy. Emotions are basically built-in shortcuts in reasoning or motivation (usually both). As such, the problem is deciding which ones to enable, which to "hard wire", and which to forbid.
E.g. child-fire: The burnt child dreads the fire.
Naive thinker encounters attractive object of class O.
Naive thinker experiences "pain" on interaction with object.
Naive thinker learns "fear interactions with objects of class O"
def. fear: Desire strongly to avoid due to expectation of injury.
So now the thinker is less naive, and it has a reasoning short-cut to use WRT objects of class O.
The real problem is that there's LOTS of specialized rules of this sort. Some of them need to be built-in because learning them the hard way would be strongly injurious. There probably aren't many like this, but the ones that are needed are very important. E.g., there was a robot in, I believe Japan, though it might have been Florida, that started disassembling itself. Whoops! And no pain sensors to warn it, either, so there wasn't any obvious way for it to learn in time not to do that.
Note that what rules need to be built-in is strongly dependent on what sensors are available.
Well, according to the latest evidence quantum effects *are* involved in photosynthesis. Not that it couldn't happen anyway, but the efficiency implies that the model using quantum effects to explain the process is the correct one.
Now I don't see any particular reason that this should translate into anything related to thoughts, but I do feel that it means that it's reasonable to consider the matter as undecided until we figure out how things really *do* work. My favorite model of sentience doesn't require, or even suggest, any quantum effects, but there are a few steps where, if quantum effects were available, it could speed things up. (OTOH, I'm not real impressed by the speed of neural processing, so even if there is a possibility of quantum effects, my bet would be that they aren't being used.)
The Turing test is sufficient, but not necessary, for sentience.
Mind you, I'm talking about the formally specified Turing test, not the bunch of half-assed imitations that people throw around as if they were equivalent. (IIRC, the Eliza program once passed a typical informal specification. I.e., it fooled a interrogator into thinking that it was a real person. It also infuriated them to the point where the person who hooked it up to the teletype nearly got fired before he convinced his boss that he had been "talking" to a machine.)
But biological machines (e.g., people) have been optimized for quick replication rather than for durability. Just as an example, the DNA code could be a lot more error-correcting than it is. But it would slow down replication too much, so it's not selected for by evolution.
OTOH, there's no essential difference between a designed biological form and a nano-machined entity. It's possible that designed biological life could be a lot more durable than one derived from mechanical basis. But not necessarily. Easy to evolve and efficient in operation are just about orthogonal axes. Once you start talking about designed biological forms, however, you leave behind the constraints of "easy to evolve", and we don't know what that design space is like. E.g., perhaps a better storage ROM than DNA is available, one that has a lower error rate, and is usable over a wider thermal range. We don't even know how much of current DNA codes are useless. We call a lot of stuff "junk DNA", but that may just mean that we haven't yet figured out what it does.
So I'm not willing to presume a particular distinction between nano-machined life forms and designed biological life forms.
Right or Left, it doesn't make much difference. They both lie freely. Allowing a "news program" to intentionally misrepresent the truth was a very bad decision. Don't know what would have been a good one, though. If you're vulnerable if you present lies, then someone will say that you lied, and probably be able to prove it. (Nobody can be accurate all the time.) I've generally given up on the media as a source of truth, and depend more on bloggers. And THEY lie. But they're much less professional about it, so the lies are usually easier to spot.
It won't just be more of the same, because the environment is changing. The prevalence of automated warehouses changes what's expected of automation. So does the initial appearance of cars that can park themselves, and the prototypes of cars that can drive themselves. Already we are starting work on embedding a model of the external world into the computer. It would be a mistake to all those things intelligent...exactly...but they make information available to a controlling intelligence that wasn't available in prior times. And just think of the competitive advantage an intelligent voice answering system would be over the current menu tree (whether it recognized your voice commands or not). So when someone learns to build one, there will be a ready market. The problem here isn't intelligence. It doesn't take much in the way of intelligence. It's building an abstract model of the external world. Building such a model DOES require intelligence, but updating such a model requires a lot less intelligence, so it only needs to be built twice. At that point a evolutionary "arms-race" will start, and development can be predicted to be rapid.
So one of the basic questions is "How intelligent a model of the world does a system require in order to out-compete a phone system menu-tree?" My assertion is "not much".
For reasons like this I think that the problem is the building of mental models, rather than intelligence per se. (And I'd assert that Bayesian reasoning is pretty close to the definition of intelligence. It's just that intelligence isn't worth all that much by itself. It needs tools to work with.)
It's one of about four or five conventional mechanisms. Generally each of the versions lead into the others, however, so there's no real conflict between them.
E.g., neural interfaces connected via a development of the Internet could yield an "overmind" capable of addressing problems that cannot now be addressed.
Or, those same neural interfaces connected to a computer via an advanced programming interface could enable the development of programs not currently possible by using feedback to stabilize one's thoughts while editing. And this could be used to...
Or molecular biology & genetic engineering could yield more intelligent people who...etc.
Or ...
Plug in whatever your favorite technological improvement is, and see how it advances the rate at which things change.
The singularity happens when the rate of change becomes too fast for us to deny that we can't keep up with it. I'd say it's well underway, but I don't expect it to arrive in full blown form until sometime after 2020, and I project 2030 as when it runs totally out of control (i.e., under it's own control ... or just uncontrolled, depending on exactly what form it takes).
If we had a government that was ethical and gave serious concern to the quality of the lives of it's citizens, then I'd be doing everything I could to slow down the arrival of the singularity. (Bar forcing civilization to collapse.) As it is, however, if we can manage to live through the transition phase, the singularity may be the only hope of survival for humanity. (My estimate of the odds isn't cheery, hovering around 40%, but my estimate if people stay in control of civilization heads towards zero within a century or two.)
We may not know the limits, but we can put bounds on them. E.g. an upper bound on the number of bits mapped by the brain is the powerset of the number of synapses in the brain (at any one time, but even though the number varies, it doesn't matter which time you pick. It's a pretty generous upper bound.)
I'm sure that someone could justify a much tighter bound, but I don't want to get into quibbles about frequency vs. pulse domain coding of neuron firings, etc., and I'm well outside my specialty.
Maybe.
I didn't read all your post, because I don't agree with your basic concept. I suspect that intelligence is rather simple, though just how simple I'm not certain. And I suspect that there are many approached to intelligence. The neural net is the only one for which we have an existence proof, but that doesn't imply that it's the only possibility.
OTOH, to interface an intelligence usefully with the real world (i.e., to enable it to understand the external reality) is probably much trickier. But that's getting extensive work all the time.
Another problem, and one that probably can't be solved until after the intelligence has a reasonable model of the external world, is language. This is separate from intelligence, but definitely acts as a facilitator for increased intelligence. (It also tends to induce massive blind spots. So there needs to be a secure model of the external world before it is introduced. People tend to introduce language too early in the process, or at least so I believe. This leads to the creation of beliefs in externals for which there is no creditable evidence. Like faeries and other such entities.) Ideally one would introduce epistemology before introducing language, but that's probably impossible. Bayesian reasoning is probably as close as we will be able to come, but it's been proven that given separate sets of priors one cannot assume that additional evidence will cause two reasoners initially different to converge. It's just the best we can do. (But it's better, i.e., more accurate, then human reasoning. OTOH, it's a lot slower. We have evolved to prefer short-cuts that lead to quick decisions over extensive evidence examination.)
So things ARE progressing rapidly towards AI. You many not notice all the changes, or even recognize them when you encounter them, but they are progressing rapidly. I still expect an AI as intelligent as a human by 2030. I'm far less certain about whether this will be desirable. That depends a lot on what it's motivational structure is. And that's something I haven't noticed much progress on...which I consider quite dangerous. (OTOH, lots of the details are being done by companies which are keeping their advanced work secret. You decide whether you trust that dynamic.)
Why?
If there's a treatment, then it might be useful to know early if you're going to get it. If there's no treatment, it only means that your health insurance is going to start looking for excuses to cancel your policy.
Students aren't in any position to reform the system, so sabotage, by one means or another, is their only reasonable response. Many teachers seem to feel equally incapable, but they have a few other choices.
(It's unfair to expect every teacher to start a Summerhill, or John Woolman, or other alternative school. But they could change jobs. Given how poorly they are normally treated, I'm surprised that they don't all quit after their first year.)
But for children...just how to you propose that they should fix the problem.
Yes.
Dynamic braking recovers much of the energy used in accelerating the car.
There are other considerations that make the answer lots more complex, but your particular point isn't a major one. (Mind you, different hybrids make different choices at many points of construction. Some of them clearly AREN'T any better than an efficient gasoline car. For others, it's not clear. Some are clearly more efficient in certain environments, and worse in others. Etc.)
E.g.: The manufacture of the car may well produce more emissions that the running of the car will produce in it's lifetime. So the best "green" statement is likely to be keeping your car running with a well tuned engine for as long as you can. In this context, the batteries of hybrids are pretty much of an unknown. Building them costs and unknown, but large, amount of pollution. How long will they last? Will they be recycled into new batteries? What's the replacement cost? What are the failure modes? (Do they just start not holding a charge?)
For now, I'll continue with a 1990's Toyota Corona. (I think that's the name, I can never keep straight between the Corona and the Corolla. I use whichever is the smaller of the two.) It's only 15 years old so far.
Basically, you're right. In the short term. It needs to become possible to maintain a closed ecology before space colonization is practical, and until then there's no place to go. (Also, don't expect it to be libertarian when it arrives. Things are going to be much too critical for that.)
This doesn't make it less important, but it does mean that expecting it right now is unreasonable. Biosphere 2 rather proved that.
Additionally, it's going to be important to improve the generation of power from solar sources. And energy storage needs to be improved. And extraction of resources from low valued ores...without using water. Etc. Lots of things need to be developed...but the thing is, most of them need to be developed anyway.
When it actually becomes possible, I expect a massive increase in interest. Even before that a permanent Lunar observatory on the far side will probably happen. That will be used to work out the final problems. (But we're aways from when that's practical.)
As to why? I don't know. Political or religious dissidents fleeing repression? Not impossible. Exporting prisoners is unlikely. I think the first colonists will need to be quite wealthy.
A space elevator is probably too ambitious for a first skyhook. Probably we should start with either the star-whale of the pin-wheel. I prefer the pinwheel.)
The pinwheel is an orbital mass with several arms that rotates sufficiently to bring the arms around. It can be in various orbits (and probably should be, to enable one pinwheel to "hand off" to another. For the lowest one, one arm rotates down into the high stratosphere, where a cargo pod is attached. (This may contain people, but it's basically unpowered.) At a certain point, the capsule is released from the arm and is flung off at a target destination. As with all sky-hooks, it's important that as much mass is lowered to the ground as is lifted. (Or it drifts out of orbit.)
The starwhale is an electromagnetic brake (or accelerator). You shoot the payload up to orbital height, but not to orbital velocity. The starwhale swallows it, and the electromagnetic brake accelerates it to orbital speed. It can then act as an accelerator to direct you to any destination you choose. (It may need to be quite long in order to use reasonable G forces for this.) The trick is, you've got to enter the electromagnetic brake properly. This can be quite tricky, especially if you're coming in from, say, Jupiter's moons. You've got a huge velocity, and you need to be properly aligned. If you miss...ugh! Both of you could be destroyed.
As I said, I prefer the pinwheel. There you don't risk destroying everything if there's a slight orbital hangup. (But you'd better have enough delta-V to be able to take another pass at the hook. But that's a much smaller problem.)