Ray Kurzweil might be right. It could very well be that Moore's law can be applied to all forms of technology, and within a couple of decades clean, renewable forms of power production will be so cheap they will have replaced all fossil fuels. Hey, it could happen. Maybe even it's not just possible, but probable. Kurweil calls it the law of accelerating returns:
Today, solar is still more expensive than fossil fuels, and in most situations it still needs subsidies or special circumstances, but the costs are coming down rapidly -- we are only a few years away from parity. And then it's going to keep coming down, and people will be gravitating towards solar, even if they don't care at all about the environment, because of the economics.So right now it's at half a percent of the world's energy. People tend to dismiss technologies when they are half a percent of the solution. But doubling every two years means it's only eight more doublings before it meets a hundred percent of the world's energy needs. So that's 16 years. We will increase our use of electricity during that period, so add another couple of doublings: In 20 years we'll be meeting all of our energy needs with solar, based on this trend which has already been under way for 20 years. (Climate Desk)
Even if the schedule is a little off, and we don't manage to curtail our fossil fuel emissions before triggering a runaway greenhouse effect, it will be still be possible to geoengineer our way back to safe levels of atmospheric carbon dioxide and methane, and cool the planet back down, and return oceanic pH to normal levels. So says Kurzweil.
But there's also at least a reasonable chance that his faith in technology is misplaced, that he hasn't anticipated some interruption that will throw his predictions out of whack, and things won't work out the way he foresees. Is it wise to base our entire strategy of dealing with the most serious public policy challenge of our time on the assumption that the cost of solar technology will continue to fall by 50% every two years? It that the kind of policy path we want to take?
Or do we want to err on the side of caution, and assume that if the business of emitting carbon as usual continues, the planet could become significantly less hospitable to civilization? Which assumption makes more sense?
I hope Kurzweil's right. Most criticism aimed at the guy tends to involve his penchant of untried longevity therapy that, while a bit on the loony side, are more ad hominem than factual, and have little to do with the fundamental question of whether we are indeed headed for an AI-bio-nano singularity or not. And on that issue, his predictions have proved remarkably accurate in the past. But I also hope that someone will offer me a six-figure, full-time job to blog about climate change. And I'm not going to bet my mortgage on that hope, either.
- Log in to post comments
More like this
There is NO SUCH THING as 'clean' or 'renewable' power.
Solar power? Solar panels use non-renewable rare earths (and so damage the environment through mining), can only be manufactured via heavy industry with all its attendant flaws, and have a finite lifespan.
Wind power? Same thing. Do you think windmills magically appear out of nowhere?
Hydro? All the major rivers in the world are already dammed (and damned) for power, and we're only beginning to understand the real impact this wholesale perversion of natural water patterns has had on the environment. (Would you believe that the Three Gorges Dam causes earthquakes? Me neither, but it's happening nonetheless.)
Nuclear power? Don't get me started.
None of this vitriol is directed at your post - you're absolutely correct that counting on future technology to correct present mistakes is profoundly foolish. It just annoys me to no end that so many environmentalists buy into the technocratic rubbish of anti-nature elitists like Kurzweil, dreaming of 'renewable' energy technologies that will let them somehow maintain their unsustainable Western consumption patterns without harming the planet. That won't happen. But the reality - that we Westerners must accept drastic population controls and a reduced standard of living, or destroy ourselves and the planet in an orgy of Malthusian annihilation - is so unpalatable that most people grasp at any renewable-energy straw (no matter how thin).
Re: Moore's Law. The big difference here is that in the computer industry there wasn't a very powerful global industry fighting against advancements in computers. The fossil fuel industry does not want to see a switch to other forms of energy, and they spend millions of dollars a year on political lobbying and on astro-turf groups to undermine funding to alternate energy research (this week it appears the GOP are following their orders quite nicely).
mad the swine has some good points. There is no "clean" energy. We just have dirty energy and either "less dirty" or "different dirty" energy. Hopefully, "less" and "different" will have more limited detrimental effects (as opposed to global...but who knows what surprises might be there).
The bottom line though is that westerners consume far too much and far too wasteful. Developing nations see what we have (through the magic of tv) and they want that too. It is unsustainable and getting worse. Magically eliminate climate as a problem and there are still half a dozen things that are going to cause wide-spread misery and chaos over the next century. We need to limit ourselves before we're "limited upon", so to speak as the house always wins. (how's that for mixing, mangling, spindling, folding, mutilating metaphors?)
It's not the technology, it's the political will to push for it... and stop subsidizing fossil fuels.
If the governments of the world took away the fossil fuel subsidies, those companies would STILL make a profit.
If that money were redirected to renewable (@#1, renewable refers to the source of fuel for the systems, not the systems. You must have never seen a refinery. I'll take the damage done my solar and wind over refineries and supertankers any day), then the US could increase the percentage of electricity provided by renewables 50% in one year (pending construction and production times, but it's a lot easier to build and install a bunch of turbines than it is to build one coal plant).
No, it really can't.
Quiz:
1. What was the air speed of the fastest commercial airline flight a decade ago?
2. What is the air speed of the fastest commercial airline flight today?
QED.
If Ray Kurzweil had been alive in the 19th century, he'd have been predicting supersonic steam locomotives within 50 years.
Ray should try reading science fiction instead of writing fantasy.
SF writers in the 1920-1960 era routinely did the same kind of extrapolation that Ray does, except that they applied it to transportation technology. I mean, look at it: in 1894, when my grandfather was born, the fastest travel in the world was a 60 mph train and most people got around no faster than a walking horse. Before he died, he watched men walk on the moon and drove at 80 mph to an airport where he took a 707 at 500 mph across the USA. Military jets were flying at more than Mach 3. Speeds had been increasing exponentially during his whole lifetime.
So obviously, in the 40 years since then, we've increased speeds as much as we did between 1930 and 1970, right?
Maybe not. Likewise with solar power.
It is true that transportation technologies don't seem to follow Kurweil's law of accelerating returns -- insofar as speeds and fuel efficiency have stagnated for decades (although the COST of transportation has fallen markedly). So fair point.
But the law does seem to apply to electronics, and I suspect Kurweil focuses on solar PV because of the similarity of the technology with computer chips -- and recent experience. So perhaps I shouldn't have summarized his theory as applying to all technologies -- just the information and electronic fields.
At least you put Malthus right there in your post so it's easy to see where you get your flawed and pessimistic misanthropic ideas.
Kurzweil is likely too optimistic about solar power but that doesn't justify your doomsday predictions or insistence that humanity cull itself down to a small group (that no doubt you envision contains you) of survivors that will drink rain water and their own urine and eat just what they can grow in a three square foot plot using their own feces as fertilizer.
Oh, and why do you limit "drastic population controls and a reduced standard of living" to "we westerners"? Has it escaped your notice that the developing world (China, India etc.) is growing economically and in population at a much faster rate than the west?
I just came from East Africa and I have seen the way people live when they practice a "low carbon" lifestyle. You couldn't take three days living the way these people live. Oh, and they are eager to attain the lifestyle you deride but of course enjoy.
Also they have exploited every inch of arable land and struggle to maintain even the basics of survival while having displaced almost all wildlife with goats, sheep, cattle and donkeys.
Almost every stream is choked with sewage and trash and all while living a "lifestyle" that you want to push on everyone. Well, everyone that survives your "drastic population controls".
Excuse me if I decline participating in your plan.
For now, and with the emphasis on the word "seems"...
The point both D.C. Sessions and myself were making is that there was a time in the fairly recent past when exactly the same sort of progression did seem to apply to transport - right up until we hit the physical engineering limits of the technologies concerned.
The usual analogy is that a falling body (in vacuum) keeps accelerating right up to the point where it hits the ground.
Did you notice how near-continuous processor speed increases and regular process shrinks started to be replaced by an increased emphasis on multiple cores and increased parallelism at some point in the last decade? It turns out electronics have engineering limits too.
Re: transportation, I believe Kurzweil uses this as an example in one of his books: nothing human made can expand exponentially indefinitely, resources will always run out. He theorizes that technology follows an S curve: a long, flat period, then exponential growth which eventually experiences exponential deceleration, then things flatline again. If you asked an owner of the first cars what speeds man would achieve within 200 years, I doubt even the bold predictors would have believed we'd surpass 100 mph. Instead, we're rocketing through space. It may be that we'll eventually turn a serious eye to interstellar travel and again break boundary after boundary in travel speeds, only to again plateau.
Re: standard of living, I highly doubt that you can place meaningful limits on human consumption. We want more food, we want more meat, we want a more comfortable life. The idea that people will willingly forgo all that, or that an authoritarian structure could force it on them forever, seems to misinterpret human nature. If we achieve a reduction in population, people will start having babies again. If we destroy industry, people will rebuild it. I'm of the opinion that the safest path is to push forward with technology with all we've got, hoping it will solve the ills of industrialization, because if we turn our back on progress somebody else will eventually try again, this time with fewer resources and a more damaged ecology.
"There is NO SUCH THING as 'clean' or 'renewable' power."
BS.
"Solar power? Solar panels use non-renewable rare earths "
Uh, those rare earths aren't used up and are, in fact, available in highly pure form for less costly reforming into new solar panels.
"Wind power? Same thing."
Yes indeed, same thing.
"Hydro? All the major rivers in the world are already dammed (and damned) for power"
Bugger me. When did they do that? In the past five minutes?!?!
"mad" indeed.
You have a nearly orgasmic hate-on for environmentalists. Just remember, for every action there is an opposite and equal reaction, acting on different bodies.
The problem with mad the swine is that some fraction of the time, zie posts sincerely, and even interestingly. Perhaps more often, zie is trolling and/or demonstrating Poe's Law. I had killfiled zir some time ago, and recently took zir out of my killfile. It may be I should put zir back in, for the sake of my peace of mind.
However, that is sustainable if there is only a fixed demand for solar panels. There will be a far greater demand for solar panels and wind turbines, meaning there will be a far greater demand for environmentally costly rare earth metals. This will almost certainly occur in spite of expected increases in efficiency of power production per unit. The only way around this is expecting some convenient substitution with more common materials.
Well, for big dams, start contracting your butt cheeks. There is a fair amount of room for low head hydro, but that still has environmental and other costs.
For now. However, those who aren't in the business (I am) miss the fact that the highest-volume process nodes are a decade or more old, and likely to stay that way. It's just too freaking expensive to do any but the most high-volume performance-at-any-cost chips on the new(est)(er) processes for a whole host of reasons.
We're headed for a time when it's just not economical (never mind physically possible) to scale chips any more. The "faster with each generation" part of Moore's Law is already breaking down for various reasons, being compensated by the "bigger and bigger" part. The "cheaper and cheaper" part is breaking down for all but the highest-volume stuff, and eventually even Intel is going to sweat the NREs if nothing else.
So Moore's Law is another "for now, but don't expect it forever" example. Solar power is limited by a number of cost factors which aren't improving, and therefore will fairly soon dominate.
Most solar panels do not use rare earth metals (which are not that rare, btw).
Also, new discoveries might totally avoid the need for rare earths in solar panels: http://www.popsci.com/science/article/2010-02/ibm-develops-higher-effic…
The main problem right now is cost: both in energy and manpower. AFAIR, it takes about 5-10 years for a solar panel to recoup the initial investment.
There's an absolute maximum to information processing based on calculation being a reduction in entropy and, even if you get away from the fact that this energy density would have the heart of a star in your PC, it is also energy density and since E=mc2, mass. So you have a limit that you will get a black hole in your PC, out of which you can retrieve NO INFORMATION whatsoever.
As to the real Moore's law, density has a problem that, even if you use individual electrons for logic, they must occupy a certain space as per uncertainty principle and the smearing of the quantum reality of the electron (Schroedinger wave equation).
Even that is a problem because you have semiconductors that have a maximum field strength before coronal breakdown and you lose your logic. A 1V logic gate 100 nm across has an electric field strength of 10^7 V/m. Enough to break down air.
And so on up to the near-micron processes we have where problems with charge containment makes it look like our current work is reaching its limit in silicon.
And lithographic printing having problems with a mask of 200nm...
"However, that is sustainable if there is only a fixed demand for solar panels."
Uh, WHISKY TANGO FOXTROT you on about?
230km on a side would be enough to power earth in 2030 with solar alone. Compare to the size of the earth.
There's plenty spare.
Rare earth elements aren't that rare.
And the winding motors for wind turbines generally don't use rare earth elements.
You are EXTREMELY clueless, yet oddly enough, strangely ADAMANT about what you know nothing about. Dunning and Kruger may have used you in a study...
"There is a fair amount of room for low head hydro, but that still has environmental and other costs."
Rather different from your earlier statement.
Still either wrong or, for the second half of that statement, FUCKALL to do with renewability of power.
"AFAIR, it takes about 5-10 years for a solar panel to recoup the initial investment."
Compare to nuclear power stations or coal fired ones.
And, as opposed to having to wait until it's complete, you can use the power generated from one of the ten thousand other solar panels yet to be built.
Try running a coal fired station with only the foundation cleared...
Problems, maybe -- but for all of that Intel is shipping one Hell of a lot of parts with channel lengths of 35 nm and below, and I'm doing an awful lot of work with 1.2 volt devices across 100 nm drawn channels.
SH*THEADS
atheistthinktank.net/thinktank/index.php?topic=9870
OH the CALAMITY!!!! OH THE DRAAAAAAMMMA!!!!!!
Think of the CHILDREN!!!!!
Think of the POLAR BEARS!!!!!
dumbass douchebags.
And for good reason. Take a lesson Star Trek Boy.
www.wattsupwiththat.com
The Singularity will never happen. The Singularity should not be relied upon to rescue the Earth from AGW because it won't.
A machine of a given intelligence can't design, build and evalutate an intelligence greater than its own.
Since the more intelligent machine will be able to do and understand things that the less intelligent machine cannot, there will be no way for the less intelligent machine to evaluate whether those things it cannot understand are actually correct and rational or wrong and insane.
The problem of the singularity is analogous to the halting problem of Turing Machines. How can one machine determine if another machine is more âintelligentâ than it is and not simply insane? That can be set up as a halting problem (but more complicated). There is no solution to the halting problem, there will be no solution to producing arbitrarily intelligent machines. It simply will never happen.