Stephen Hawking: Off Earth by 2110?

[Junghalli]

What, in practice, is the difference?

None, because learning capacity is an integral part of intelligence.

All things being equal speeding up should increase effective learning capacity. The person can study more in the same amount of time. It would help if we knew what definition of "moron" was being used here: uneducated, low IQ, wilfully ignorant? The first should definitely be improved by speeding up, the latter probably would be somewhat, the last probably wouldn't.

If we actually apply computing principles, the smart thing is to have it automatically underclock itself when not in use, like computers do. When thinking (high CPU use), the clock rate automatically, and near instantly, is boosted. When not thinking (idle cpu), the clock rate instantly and automatically drops.

The sleep issue would probably solve itself anyway. If all the brain processes are happening 7X faster then logically so will the sleep cycles.

[HMS Conqueror]

Apparently, you live in a world where all problems can be solved with enough spending and wishful thinking. I used to live there too. And then I turned twelve. Economic growth is only exponential if you still have resources to spend to sustain that growth. You can only flog the planet for so many resources before the cost of extracting that next unit of resources becomes more than you can readily afford.

Ah, it was the opposite for me: when I was twelve I thought resource depletion &c. was a serious problem, but now I am at university I no longer do. There is in fact no clear link between resource consumption and economic growth. The invention of the silicon transistor, for instance, made computers better and cheaper and less consumptive of resources. The only potentially fundamental limit that exists is imposed by energy production, and that is far in excess of anything we will reach any time soon.

And yes, we're liable to bump right up against the edge of scientific knowledge in a century or two.

I can only lol. How can you possibly justify such a prediction?

That fact should be profoundly worrying, as it suggests that there aren't going to be any magic-tech solutions to get us out of the corner we're painting ourselves into. I suppose the name "Malthus" means nothing to you.

Malthus is famous for being wrong, you know. If he wasn't, modern society would be impossible and humanity would have already experienced the predicted apocalypse.

lol, there is no conceivable shortage of energy on earth for millions of years.

Proof?

This graph is a bit of a mess but I found it this morning (unrelated to this forum), so I might as well use it:

Mind you, there is the sun, but I've already established that we've blown the equivalent of millions of years of solar energy input in the form of fossil fuels.

The implication is that because fossil fuels take millions of years to produce that they store 100% of incident solar energy for millions of years. In fact the conversion efficiency was a minute fraction of a %. That's not to say that running out of fossil fuels won't be bad: they're super-awesome in general, being both cheap and convenient for use. But it will be more like a few percent (not ppt) reduced economic growth for a couple of years or decades, not the civilisation implosion being peddled by the peak oil movement.

Allegedly, we can sustain current levels of energy usage through nuclear power for many megayears . . . assuming the widespread adoption of breeder reactors, and the profitable large-scale extraction of uranium from seawater. All of this assumes that we'll spend the few trillion dollars needed to convert our petrochemical energy infrastructure into a nuclear/solar one before we hammer the planet's petrochemical reserves to the point where we're spending huge amounts of money just to keep up with our energy demands.

And don't forget that energy is only one part of the puzzle. You also have to feed all those people, and they have to live somewhere. And they have to do so while dealing with the legacy of climate upset that we'll be leaving them.

Those are already solved problems. 90bn, maybe that's going to be a big issue, but not 9. That's only a 50% increase.

Climate change is an issue, I agree, but not an insoluble one. Especially not with the nuclear/synfuel economy.

The only issue is the cost of generation, which economic growth takes care of over time.

And what is going to fuel this endless economic growth? Fairy dust and unicorn shit?

Capital accumulation and technological progression - just as in the past.

The reason we use the cheap-and-nasty methods first is that we have less money and technological capability in the present (whenever that may be) than in the future.

Oh yes. We've moved from cheap-and-nasty methods to expensive-and-nastier methods. And that's just to keep up with growth in demand. Yet I don't see any multi-trillion dollar projects on the horizon to convert our petrochemical lifestyle to a nuclear/solar one. Do you?

It's not yet viable (nuclear that is, solar won't be viable for centuries), but it's off only by single or low double digit percent, not some huge amount. And that's with extensive political damage to the nuclear industry. If you removed that, and added the externality cost of AGW to fossil fuels, we would already have 80-90% nuclear electricity generation, as they have in France.

[Formless]

All things being equal speeding up should increase effective learning capacity. The person can study more in the same amount of time. It would help if we knew what definition of "moron" was being used here: uneducated, low IQ, wilfully ignorant? The first should definitely be improved by speeding up, the latter probably would be somewhat, the last probably wouldn't.

The very fact that the last is a possibility demonstrates that there is more to intelligence than just clock cycles. That's my whole point. There are multiple functions that must be accounted for that combined measure intelligence. Learning capability, memory, problem solving, honesty/open-mindedness (as a practical measure though not necessarily a innate property of the system), and yes speed of thinking. And that is just to name a few.

The sleep issue would probably solve itself anyway. If all the brain processes are happening 7X faster then logically so will the sleep cycles.

But then we get back to the problem of boredom and the fact that the person is essentially alienated from any kind of environment he can relate to. You've essentially put him in a prison he cannot ever escape from. Why would you ever want to do such a thing, when you could just design an AI from the ground up, using operating principals more suited to what you want?

[Sarevok]

I think we should not ignore the biological and chemical buildup of the brain when talking about simulating one. The stereotypical "moron" being simulated could be result of many very complex factors. This is why I think any arguement on superhuman intelligence based on simulating people is on very shaky ground to begin with.

[HMS Conqueror]

Then it should never be done. However, since economic growth seems to be exponential I would not be so intimidated by orders of magnitude.

It only seems to be exponential because that's our recent experience. There have also been times when the economic engines reversed.

For an instructive example see Easter Island. It's a beautiful example of what happens when overpopulation crashes hard. Just because the 20th Century was all progress (actually, it wasn't until around 1945 in the US, and later than that in the rest of the world) doesn't mean the 21st Century will be.

Of course humanity won't implode any time soon. We could keep shambling along like this for, at least, another century.

If by shambling on you mean reaching the pinnacle of material prosperity, scientific knowledge and population size - all still increasing - then I see no problem for humanity's survival.

You know, Ancient Rome was pretty damn snazzy, too - then the Dark Ages came along. Gee, that really sucked. How about the collapse of the Mayans?

Civilizations can and do fall. Typically, what's left behind never re-achieves to the extent of the Big Golden Age.

Easter Island's economy imploded due to political institutions that caused rampant tribalist looting of all private property. Whatever you might have to say about the Obamessiah, and I'm as opposed to him as the next man, he isn't that bad. Nor are we heading for late Roman Empire-style economic despotism. Ok, so it is theoretically possible that the West will be overthrown by some Maoist dictatorship that will destroy the living standards, but it isn't actually going to happen. Come on, even the hard left in the West basically accepts market economics to a great enough extent to avoid that sort of thing.

Not to mention our current civilization runs on the Magic Potion known as PETROLEUM - which even the most optimistic admit is a FINITE resource. We're burning through it at a breath-taking rate. The best estimates I've seen for alternative sources would give us, maybe 30% of our current energy output, so cut to 1/3 all your driving, the lights in your house, your household appliance use... and that's the BEST scenario. Once it's gone a lot of the "Green Revolution" in agriculture comes to a screeching halt and we go back to 19th Century crop yields (if we're lucky - in some places the soil fertility is pretty much destroyed, so yields would be even less). People will starve by the tens of millions (if we're lucky, it could be worse). Much of our medical advances in regards to disease are becoming useless because the germs keep evolving. All it takes is one virulent pathogen to get loose and we could lose a LOT of people. Lose enough and both the economy and technology go into reverse.

That's totally leaving aside something like Yellowstone having a major eruption - which couldn't do a damn thing about. It would destroy North American agriculture and economy for several years at best - again, tens of millions starve, weather is fucked up, not only does science and technology stop advancing they might even reverse...

There are actually a LOT of threats to both our civilization and our species.

In fact, for long-term health of both the planet and our species the best "solution" I can see is losing about 75% of the current population (actually, we could happily get by with less than 1 billion) and there's much, much more to go around for everyone be it oil, land, food.... except for the nasty, nasty detail that anything that takes out that many people would be absolute hell on Earth for everyone who goes through it, be it plague, war, famine, or natural disaster. And, depending on the mechanism, it might take a LONG time to rebuild to even 20th Century levels of ability.

So, do us all a favor - study history a little better.

Rather, petroleum use will be reduced and replaced as the price rises, causing somewhat suppressed economic growth for a while. Oil prices already at least doubled in the last 50 years, without much noticeable effect.

lol, there is no conceivable shortage of energy on earth for millions of years.

You fucking moron - you miss the point. NOTHING gives as high a return on investment as petroleum. NOTHING. At least nothing we currently have. Coal might be a close second, but it's polluting as all hell. Nuclear requires a certain level of high tech to build and maintain. Solar - well, plants really aren't all that efficient, there's just a LOT of them and our artificial solar isn't that much better (if it even matches). Wind - not reliable 24/7. Geothermal - limited to certain locations. Hydrothermal - most of them best locations already built up, and there's an environmental cost to using it. Sure, there will always, in a sense, be available energy but nothing we have is as concentrated, portable, and adaptable as oil and its derivatives.

Petroleum arguable is a negative return on investment (in terms of energy) depending on which figures you believe. That isn't the point of it: petroleum is good because it's both energy dense and a liquid, so is convenient for use in vehicles. It's already not used in bulk electricity generation, because it's too expensive.

Nuclear requires high tech for sure, if you've been in a coma since the 1940s.

[Junghalli]

The very fact that the last is a possibility demonstrates that there is more to intelligence than just clock cycles. That's my whole point. There are multiple functions that must be accounted for that combined measure intelligence. Learning capability, memory, problem solving, honesty/open-mindedness (as a practical measure though not necessarily a innate property of the system), and yes speed of thinking. And that is just to name a few.

I'll happily admit there's more to intelligence than raw processing capacity (otherwise you could make a chimp superhuman by speeding it up enough, which I very much doubt is possible), but if you have a remotely capable mind increasing processing capacity should give you some improvement. If you take a good engineer and speed him up so he effectively has years to work while a human one would have hours, then I think you could say with some accuracy you've created a superhuman engineer. He can solve problems much faster than any human engineer. So minds on par with average competent humans + ability to speed up their clock speed drastically = effectively superhuman minds.

Incidentally, another transformative possibility with human-level minds in silicon is if they could be copied like any other software. Simply being able to mass-manufacture minds would be radically transformative, even if you leave out superintelligence. Imagine if, say, instead of training thousands of engineers you could simply take your best couple of them and copy them thousands of times. This would potentially give you a lot of the same benefits as a superintelligence-driven intelligence explosion (e.g. progress on cancer cures might happen a lot faster if we had millions of copies of our best researchers working on it).

But then we get back to the problem of boredom and the fact that the person is essentially alienated from any kind of environment he can relate to. You've essentially put him in a prison he cannot ever escape from. Why would you ever want to do such a thing, when you could just design an AI from the ground up, using operating principals more suited to what you want?

The easiest solution would be to slow his clock speed back down to normal during free time and accelerate it during work and sleep time. Of course if we're assuming AI with being able to do some degree of custom-designed cognition you could probably do a lot better than a speeded up human, it was just an example to show the easiest and simplest way you could go from a human-level mind to an effectively superhuman mind.

[ThomasP]

RE: overclocked brain emulations

Is there any particular reason that speeding up specific cognitive functions or elements of intelligence couldn't happen independently? If you could speed up say working memory and the regions responsible for abstract reasoning, language, etc, while leaving normal perception of time and consciousness alone, you could end up with an emulation that can process and absorb information much faster than normal person but with no real difference in how time passes. You'd at least avoid that problem of having a Palin voter trapped in an infinite library.

Granted it may just be easier to drop the whole thing in a virtual machine and run it at 16X, but I'm curious as to whether or not that would be an option. I could see it requiring some in-depth neurological knowledge that the brain-em scenario intentionally leaves in the black box, but if emulations are being run as different modules it might be doable.

[Formless]

What you are describing sounds an awful lot like racing thoughts. Just something to keep in mind-- the results might look (and feel) a lot like a manic episode.

[Singular Intellect]

If nothing else, it can design better and better circuit boards, while optimizing its own code. Besides, a properly designed AI will have vastly different thought processes to those a human; we've got a bunch of clunky relics of our evolutionary history distorting our worldview, and an AI wouldn't have those unless we intentionally programmed them in.

And suppose there are hard physical limitations to how sophisticated computer hardware may be made?

Of course there will be limitations, but any notion that we're even close to them is ridiculous.

According to what I've heard, computer physicists claim that if they could make molecular/atom based computers, which they are in fact working on, a sugar cubed sized computer would have more processing power than the entire planet's current computer resources combined and be monstrously efficient in power consumption to boot. And this doesn't even delve into quantum mechanics. Can't recall the video source I watched that on though.

[Setzer]

I agree with Hawking; I think his analysis is absolutely correct with one exception. It's already too late. We've missed our chance. Had we gone full-bore for space exploration from the 1960s onwards, then we had a chance but that never happened. Now even with a full-bore effort, that missed fifty years will be critical. I think this is probably the explanation behind the Drake Equations Dilemma. There is a very fine window between the technology developing to take a species into space and safety and bringing about their own extinction. That window is so fine that the overwhelming majority of intelligent species miss their chance and slide gracefully or not-so-gracefully into extinction.

This is the most depressing thing I've heard since I read what Josef Fritzl was doing in his basement. When you say we've missed our chance, you mean what? That we won't be able to develop the technology needed to settle other worlds before running out of resources on Earth? That humanity, as a species, is doomed? That we've already failed natural selection?

[Pelranius]

I think it would be more accurate to say that we (being America and Russia) have missed out on the chance to take the lead in spaceflight and development. And to look at it harshly, a global population of one billion could possibly sustain a large scale space program with some levels of advanced automated industry (though that begs the question of what happened to the other 5-6 billion who didn't make it through the resource crunch, if saltwater hydroponics, mass scale nuclear power and underwater mining don't take off).

[Guardsman Bass]

I agree with Hawking; I think his analysis is absolutely correct with one exception. It's already too late. We've missed our chance. Had we gone full-bore for space exploration from the 1960s onwards, then we had a chance but that never happened. Now even with a full-bore effort, that missed fifty years will be critical. I think this is probably the explanation behind the Drake Equations Dilemma. There is a very fine window between the technology developing to take a species into space and safety and bringing about their own extinction. That window is so fine that the overwhelming majority of intelligent species miss their chance and slide gracefully or not-so-gracefully into extinction.

This is the most depressing thing I've heard since I read what Josef Fritzl was doing in his basement. When you say we've missed our chance, you mean what? That we won't be able to develop the technology needed to settle other worlds before running out of resources on Earth? That humanity, as a species, is doomed? That we've already failed natural selection?

Stuart has a very healthy fear of biological weaponry, not without reason*. It's likely only going to get easier and easier to make that type of stuff as the century goes on, and having somewhat self-sustaining off-world colonies is one way to avoid having a biowar wipe out your civilization.

Beyond that, though, interest and investment in manned space travel and colonization beyond LEO seems to have mostly gone down-hill from the 1960s. That means extinction in the long run, even if we don't wipe ourselves out.

* I mentioned earlier that I'm not quite as pessimistic on this, though. The Soviets managed to make some nasty stuff, but they also had an extensive infrastructure with hundreds of scientists working on it (read the book Biohazard by Ken Alibek).

[Broomstick]

Let's get a grip here - there human race WILL go extinct at some point. No species endures forever. One day, no matter what, H. sapiens be gone. The real questions are:

1) are we going to be short lived or long lived as a species?

and

2) are we going to leave any descendants in one or several new species?

[The Romulan Republic]

I agree with Hawking; I think his analysis is absolutely correct with one exception. It's already too late. We've missed our chance. Had we gone full-bore for space exploration from the 1960s onwards, then we had a chance but that never happened. Now even with a full-bore effort, that missed fifty years will be critical. I think this is probably the explanation behind the Drake Equations Dilemma. There is a very fine window between the technology developing to take a species into space and safety and bringing about their own extinction. That window is so fine that the overwhelming majority of intelligent species miss their chance and slide gracefully or not-so-gracefully into extinction.

I really hate this kind of attitude. If I understand your position correctly, its basically "humanity is doomed whatever we do." The logical extension of which, for a lot of people, is "why bother trying to make things better?"

[Pelranius]

What about deepsea underwater colonies as a measure of last resort against biological weapons?

Stuart had them in the TBOverse, IIRC.

[J]

Ah, it was the opposite for me: when I was twelve I thought resource depletion &c. was a serious problem, but now I am at university I no longer do. There is in fact no clear link between resource consumption and economic growth. The invention of the silicon transistor, for instance, made computers better and cheaper and less consumptive of resources. The only potentially fundamental limit that exists is imposed by energy production, and that is far in excess of anything we will reach any time soon.

Really?

World GDP

World energy consumption

There's an interesting trend which shows up when one plays with the GDP/capita and energy consumption/capita charts on Google; mature 1st world nations such as Canada, Germany, and the US have stabilized their per capita energy consumption for the past 20-30 years, industrializing nations such as China and others in Asia have seen per capita energy use rise significantly, lastly, the poorest nations have been flatlined during this time period.

The GDP per capita charts show large gains in 1st world nations which indicates in efficiency, though I'd argue that a large portion of the gain was due to Enron style accounting and the abuse of financial instruments & credit. The industrializing nations are also seeing large GDP gains to go with their energy consumption, the gains in efficiency (if any, depending on the country) aren't enough to flatten out their per capita energy use. The poor countries are flatlined here as well, if not declining.

Put it together and we have a problem; most of the world's population lives in those rapidly industrializing nations, the ones with growing populations and rising per capita energy consumption. Efficiency is not going to save us until they're on a mature 1st world level. If they were to have the same energy consumption as a modern country like oh, Germany or France, it would require the construction of several thousand large nuclear generating stations over the next few decades.

And what is going to fuel this endless economic growth? Fairy dust and unicorn shit?

Capital accumulation and technological progression - just as in the past.

What capital accumulation? An overwhelming amount of the "capital" in existence today has around the same value as Monopoly money, with roughly the same real world purchasing power should the holder attempt to use it. A $10 billion CDO is not capital nor are non-agency MBSs, yet both are claimed as such by their holders.

Petroleum arguable is a negative return on investment (in terms of energy) depending on which figures you believe. That isn't the point of it: petroleum is good because it's both energy dense and a liquid, so is convenient for use in vehicles. It's already not used in bulk electricity generation, because it's too expensive.

Only for oil shale, possibly for some types of tar sands and scavenging the last bits of oil out of depleted reservoirs. All other forms of oil production still have a positive EROEI, generally 5:1 or greater.

[Stuart]

When you say we've missed our chance, you mean what? That we won't be able to develop the technology needed to settle other worlds before running out of resources on Earth? That humanity, as a species, is doomed? That we've already failed natural selection?

Basically yes. I would be very surprised if humanity makes it to the end of this century. Humanity is doomed although what will be the coup de grace is a good question. Trapped on this one planet, we're as vulnerable as hell to anything the universe can throw at us or we can throw at ourselves. I think natural selection for an intelligeny species is one long string of choices and to survive it has to get every one of those choices right. We were doing pretty well up to the 1970s and then we blew it by retreating from space. Now, the big question is whether there's an "escape route" by which we can get back on to the winning track.

I really hate this kind of attitude. If I understand your position correctly, its basically "humanity is doomed whatever we do."

Not quite; I think we missed our chance and now it is too late to save ourselves before the hammer falls. Say again, something will get us whether it is a rock falling from space, energy and raw material depletion, some lunatic letting an engineered bioweapon loose (my personal bet as the most likely) or (insert disaster). We're trapped on this one floating rock and there's no back-up.

The logical extension of which, for a lot of people, is "why bother trying to make things better?"

Because I might be wrong. There may be a get-out clause that we're overlooking. There may be a deus ex machine like an alien arrival that changes everything. I would assign either of those a low probability but I have the humility to admit that I might be wrong. You're entirely welcome to dance on my grave if I am wrong (just be careful to check for a sign saying 'beware landmines' before you do). We can't know that there isn't a save-the-game route so we keep trying to find one until the hammer drops. But, realism based on current knowledge levels doesn't look very hopeful. The odds are against us but then we're the only game in town. So we keep trying.

[Temujin]

I have to say I share Stuart's pessimism; although I'm a bit more optimistic (in two ways), but not much.

First, I'm convinced that the window hasn't closed completely, and that some technologies may still offer a us a way out. Even if it means a brute force approach by using nuclear powered rockets; in that case the environment will be probably be pretty well fucked by then anyway.

Second, unless something major happens (i.e, Dino killer asteroid), I don't think humanity will go out completely with a bang. More likely we'll collapse back to a preindustrial age and be effectively stuck there.

[Pelranius]

Second, unless something major happens (i.e, Dino killer asteroid), I don't think humanity will go out completely with a bang. More likely we'll collapse back to a preindustrial age and be effectively stuck there.

I doubt that we'll be eternally stuck in a preindustrial age after a global collapse. What worries me is that an industrial collapse will kill most of the world's population (6 out of 7 might be very optimistic). Sure, the survivors can probably eventually build interplanetary and interstellar colonies, but what about the 6 billion poor saps who became kaput?

[Temujin]

Well, my rationale is based on the idea that if we lose enough knowledge and technology to put us back in a preindustrial age, there likely won't be enough cheap and easy to acquire resources to allow for industrial revolution 2.0. If we keep enough knowledge and functional technology, even if its only in a few countries, than we can try to work around that limitation by using alternatives.

[GrandMasterTerwynn]

Second, unless something major happens (i.e, Dino killer asteroid), I don't think humanity will go out completely with a bang. More likely we'll collapse back to a preindustrial age and be effectively stuck there.

I doubt that we'll be eternally stuck in a preindustrial age after a global collapse. What worries me is that an industrial collapse will kill most of the world's population (6 out of 7 might be very optimistic). Sure, the survivors can probably eventually build interplanetary and interstellar colonies, but what about the 6 billion poor saps who became kaput?

No. Once technological civilization goes, that's it. We used up all the easily-extractable, highly-concentrated energy sources that we used to kick-start our own industrial revolution. All the surface mine-able coal was tapped out years ago. Coal now comes from deep shafts, or mountaintop removal. The only light, sweet crude oil left in the world is in deep reservoirs requiring complicated drilling technology to reach. The same goes for natural gas. Anything more esoteric requires a mass of working high industry to bootstrap off of. If civilization collapses, we're restricted to doing work by burning biomass (either by directly setting it on fire, or feeding it to slaves and beasts of burden.)

If technological civilization falls apart, the energy sources available to its' descendants would be enough to sustain (at best) 16th or 17th century lifestyle with the occasional bursts of 18th or 19th century activity as forests wax and wane and civilizations stockpile the charcoal needed to fire large-scale steel production/recycling.

[Junghalli]

If technological civilization falls apart, the energy sources available to its' descendants would be enough to sustain (at best) 16th or 17th century lifestyle with the occasional bursts of 18th or 19th century activity as forests wax and wane and civilizations stockpile the charcoal needed to fire large-scale steel production/recycling.

Theoretically, could a farsighted and dedicated government get back to the point of being able to invent nuclear power and spacecraft by stockpiling resources and carefully rationing them to the bare minimum necessary to sustain a high technology base?

I know it isn't plausible with real humans, but if we postulate a population of sufficiently altruistic and farsighted transhumans or whatever could it theoretically be done?

[Sarevok]

What about hydro-power and wind power ? A Earth with far less humans than today could try to jump to 20th century tech level using these two.

[Darth Yan]

apparently a big asteroid like the one in armageddon will be here in 2037. That will kill us.

[PaperJack]

apparently a big asteroid like the one in armageddon will be here in 2037. That will kill us.

you forgot about the 1-in-6,250 chance of it actually hitting the planet (99.98% probability of missing)