Futurist and inventor, Ray Kurzweil, predicts that advances in nanotechnology will result in solar energy powering the world in just 16 years. The Big Think website recently interviewed him, and here is part of the editors’ summary:
[H]ow far away is solar from meeting 100% of the world’s energy needs? Eight doublings, says Kurzweil, which will take just 16 years. And supply is not an issue either, he adds: “After we double eight more times and we’re meeting all of the world’s energy needs through solar, we’ll be using 1 part in 10,000 of the sunlight that falls on the earth. And we could put efficient solar farms on a few percent of the unused deserts of the world and meet all of our energy needs.”
So Kurzweil is an optimist. But should we believe him?
I say no for two simple reasons:
- He has no specific expertise in nanotechnology or solar technology; and
- it’s not obvious that any experts in the fields that he is making predictions in are similarly optimistic.
So why does he get so much attention with his pronouncements? I think it’s (at least in part) because people are desperate for hopeful news about the world and Kurzweil has the vague cache of being an inventor and futurist with some megatrend predictions under his belt that have largely come to pass.
But what support does he give for this recent claim? A pretty lame one: an extrapolation based on Moore’s Law. Here are the editors at Big Think again:
Just like computer processing speed—which doubles every 18 months in accordance with Moore’s law—the nanotechnology that drives innovations in solar power progresses exponentially, he says.
During his latest Big Think interview, Kurweil explained:
“Solar panels are coming down dramatically in cost per watt. And as a result of that, the total amount of solar energy is growing, not linearly, but exponentially. It’s doubling every 2 years and has been for 20 years. And again, it’s a very smooth curve. There’s all these arguments, subsidies and political battles and companies going bankrupt, they’re raising billions of dollars, but behind all that chaos is this very smooth progression.”
Why there would be any connection between the increases in computer processing speeds and nanotechnology advances (and how this relates directly to solar cells) is not explained.
Kurzweil is fun to read, but I think he’s mostly selling snake oil.
Prometheus Unbound 
Kurzweil is full of hot air. He’s been talking things up for years if not decades. He appears to be a passionate and excitable futurist who speaks well to serious audiences. Despite having a track record of recurrent, over-eager predictions which are nothing more than over-imaginative day dreams – a science or tech expert’s make-believe game, if you will – these universities and conferences keep inviting him back.
He’s made predictions in biotech and computer science that are clearly baloney (or if not false, at least he is overselling the rate at which actual scientific progress really happens). I have no doubt that this is more baloney.
In his 2005 TED Talk, “How technology’s accelerating power will transform us”, he said we will use respirocytes in our lungs to do an olympic sprint for 15 minutes on one breath or sit at the bottom of the pool for 15 minutes. This may happen but he gives no indication of how long this will take or how realistic this upgrade is. Apparently he has even predicted human immortality – due to nanotechnology – in 20 years!
One reason why i think people like Kurzweil is because he provides a materialistic, science-based utopian vision; this replaces the traditional, religious eschatology. Rather than becoming one with God, Kurzweil would have us become one with machines.
I know how much we might wish for Kurzweil to be telling the truth about all of this but it is very important to take a step back, disengage our desires, and ask, “what evidence is there that he is correct?”
but its not utopian; hes not saying that super technology is a good thing, he’s just saying it will exist. and considering the steps we’ve taken in the past 50 years – this forum, for example – theres no reason to dismiss the idea
Hi, Liggamite, how is his concept of the Singularity not utopian? I’d be interested to hear your thoughts on this.
“utopian” is defined as founded upon or involving idealized perfection. So yes, his idea is talking about a sort of extraordinary perfection, but I’m not sure you can call it idealized. Technological progress has caused death in the past, and with the kind of exponential progress he’s suggesting, it’s not that far off that our civilization would be destroyed. His ideas are futuristic for sure, and hopeful (he must be eager for the “singularity”) but I wouldn’t say that he’s dreaming of a perfect world, just an incredible one.
the world was amazed by the atomic bomb, for example. ‘How extraordinary, what an achievement by humanity.’ Yet, it has killed millions of people. Technologies can be “perfect” without being benevolent
santitafarella, there might not be any connection between computer processor doubling and solar panels at all. That is not what Kurzweil says which is why you fail to see a connection. What he is saying is all technology begins to fit the exponential growth model. Computers have stayed consist on their curve for over 100 years. We can see Solar Panels have also found their curve and its been consistent for 20 years. Like any good marksman knows you aim ahead of a moving target to hit it. He follows the curve into the future and can say, if solar panels stay on their exponential growth curve for even have the time computers did, we will meet the worlds demand for power in 16 years. No other technology matters because as we have seen all technologies stays on their own exponential curve.
Iain, you question the truth of what Kurzweil says. It a prediction based on mathematics. It would be like me predicting the chance of getting a 21 in blackjack and asking how much truth there is in it. It is 100% truth based on looking at the past evolution of the technology. He simply says solar panel technology has been able to provide twice the power per dollar every two years. It has stayed consistent for 20 years so one could predict if it continues to follow this exponential growth model it will meet the worlds demand for power in 16 years in a way that is affordable. Might not happen but the statement is 100% truth.
The reason he has been talking about this for decades and people continue to listen is simply because what he says is happening and is verifiable. Its weird that technology has got to the point where you can carry literally 1000 of albums of music with you on a device just a few square inches, movie, books, you name it, and he suggested this may happen when 4 MBs of RAM cost $600 (over $100 a MB not GB). I remember listen to him talk about this. You ask for evidence? You are so immersed in the evidence you don’t even see it. By the way if you pay attention to Watson in Jeopardy, that is actually a little ahead of schedule. But who’s keeping track of the evidence?
i agree. the things we’ve accomplished in the 20th century just makes me wonder what extraordinary things are coming. Mindblowing things. but were not made to think about that stuff
Saying that Kurzweil is simply neutrally describing an exponential curve and thus his statements are 100% truth is misleading. I read somewhere online that humans are getting taller by 10-30mm every ten years. So I could therefore “truthfully” say, based on looking at past human biology and the increase of human height, that if this stays consistent in the future then we will be up to 4.75m tall by the year 3011.
Well, that’s true. It’s also tautologous when you think about it. What is important, however, is whether we have a good reason to think this trend will continue or whether there are other constraints (in this case, genetic, material/physical, and nutritional) that might make this statistical statement unrealistic.
I’ve spoken to people who work in the solar panel industry and even they are not very optimistic about the use of solar panels globally as an alternative energy scheme. I’d love it if they were, and hope Kurzweil is right, but am not convinced he is. I wonder, what does Kurzweil know or assume that these industry folk don’t? Does he know something about solar panels that will benefit us all in the future? What constraints does he know about in the industry that he is confident that we will overcome in order to continue an exponential growth into the future?
As far as it goes with computers. They have improved nicely for a while, this is true, but recently I’ve noticed that we seem to be making gains mostly by doubling cores etc rather than by making a genuine breakthrough in the underlying technology. Quantum computing will allow for vast gains to be made but currently this technology is very much in the “research phase” and even when it does work we are dealing with a very, very limited number of operations per second (did I mention, VERY limited?). If we want computers to be smaller, cheaper, and faster then we can’t simply keep doubling up… we need an actual paradigmatic breakthrough. These kind of breakthroughs can’t be rushed. They simply need to be slowly and reliably researched until they are achieved. Kurzweil won’t make this happen by wishing it to be so. I’m a keen and avid friend of science and technology and I know we will do great things in the future but I just think Kurzweil seems to make predictions that things will happen much sooner than is usually realistic in the hard-won field of scientific advancements.
I seem to recall Kurzweil giving a talk in the 90s where he made some ten-year predictions about the technological advancement of human biology. I noted this a few years ago because I watched it AFTER the date he predicted and realised he was wrong. I’d love to find the video to show you (if I find it, I’ll post it). He’s not a magic man or a science prophet, he’s just a smart guy who is sometimes right and sometimes wrong.
p.s. where’s my jetpack?
Hello, like you blog, just the kind that I like to follow..
His predictions aren’t based purely on the observed results, but on the underlying causes. From what I have read of Kurzweil I have gotten the impression that he is basing his predictions on a LOT of statistics and market dynamics research rather than just going “Hey, this cpu die is twice as dense as the last one I got, surely I could write books about this!”
I’d say your human height comparison is a bit to simplified to be a good argument..
And Moores law isn’t something that will last forever; it lasts for this particular technology. I’m guessing it will not be applicable in 10-20 years due to to shifts to completely different architectures.
You are correct in that not much has changed in computing technology for a long time. This is kind of what I think the “singularity” business is all about. We are making slow progress, and most innovations are superficial. However at one point not too far away we will have available such huge amounts of processing power that all we need to unlock a “deep thought” style computer is the right software.
Currently we are building all our improvements mostly on hardware; the processors are getting so much faster and the architectures so much more compact that it is easy to get a big boost in performance just by upgrading a machine to the latest hardware. When the hardware advancements start to plateau more attention should be turned to software and suddenly a lot of very inventive AI should pop up. We are already building the infrastructure for knowledge sharing that will be needed (think google) and all that is missing is the right intelligence.
All it would take is ONE single AI entity gaining pseudo-sentience (self-aware on it’s own terms, not ours.) and starting to handle abstract concepts the way we do and the immediate jump in technological capability of the human race would be beyond the dreams of most sci-fi writers..
Excuse the poor comparison but:
Imagine a human being trying to write a program to simultaneously manage the various human-built power generators and reserves of the entire earth. Now imagine that human being fully aware of the state of each object in near real-time. Furthermore, imagine him being fully immersed in all human knowledge on this subject; he is simultaneously immersed in power grid design, programming, system management and the science involved in each power generator. Now imagine him having the kind of computing power behind him that could calculate PI down to the last decimal a thousand times over each second……
About the jetpack; unfortunately this technology just doesn’t seem practical. I’m sure it could be figured out in a few years if there was marketability for it, but there really isn’t.
If your really want one, maybe you should give this guy a call;
I don’t really have any definitive answers, but I do have imagination and sufficient understanding to be pretty damned impressed by some of the stuff that’s going on in the world right now 😮
yes, thats true..it just takes one greater-than-human computer and unimaginable things happen. except, theres no such thing as AI. either anything can be life, or nothing can
Well that’s a matter of definition. I believe the most common definition of AI demands that it
#Be self aware
#Be capable to understand and function in a human-dominated society
I’d agree about life, in my opinion there is nothing making us humans more “alive” than other creatures or mindless automatons that we think of; we just have a lot more high-end features.
However, we have not yet managed to create any artificial entity which is truly independently self-aware. Obviously that’s a silly distinction to make in a way, we humans didn’t magically get our sentience- we evolved into it. The sentience of an AI would be similarly derived from it’s base code.
I guess my idea of what defines an actual AI is that is does not have to be maintained by humans to continue it’s existence and reproduction. I would want an AI to be able to:
#Manufacture the hardware necessary to host it
#Update it’s hardware as well as non-base code
#Inject it’s base-code (DNA) into an empty system and have it evolve from that into something that has the same features and functions as the original.
I also believe that we as humans should, when the time comes, embrace the possibilities of this technology and enhance ourselves. Otherwise we will just have created an intellect more powerful than ourselves and we will have to either constrict it severely just be able to coexist or not allow it to exist at all, for fear of it becoming the dominant life-form in our available space. And I do not believe we can program some magical “rules” into the core of such a being to keep it in check.
Anyways, AI isn’t really my primary concern. Controlled evolution is.
“I’d agree about life, in my opinion there is nothing making us humans more “alive” than other creatures or mindless automatons that we think of; we just have a lot more high-end features.”
Exactly, and thats why it angers me when people say humans are so special and all else is dumb. I love our species, but in the next few years were going to create “artificial” intelligence that’s more “alive” than we are. There’s no such thing as “natural” life
The most positive thing on energy for the short to intermediate run is increased supply of natural gas. I’m no expert at all on solar technology, but people have talked about for along time. As a replacement for incumbent technology on a large scale any time soon (next 30-50 years), my only slightly informed opinion is no.
It’s like Jung claimed, religion is in our human nature. That’s why atheists make up these inferior pseudo-religions as a substitute. You can’t run away from yourself.
That said, who knows? Maybe solar power really is the solution. Although once we have a strong source of energy our consumption will also increase. These things are impossible to predict. I sure wish it to be true though, so we could get rid of the oil dependence.
its not optimism, its analytical realism. i think he’s a genius. his trends have been totally accurate for the increasing ability of technology over the past 200 years. youve got to remember, no one believes anything until it actually happens. We weren’t gonna land on the moon. What are cell phones? How could anyone bomb the twin towers. like kurzweil said, our minds aren’t built to think the impossible. they’re built to keep us alive. I say he’ll be correct
What precisely are you saying he’s going to be correct about? I ask for clarification because in order to know whether he is correct we need to work with a specific, clear, and well defined prediction.
Iain, read his books. Like many sceptics, you appear to be aware of the conclusions but not the detail. Take the Human Genome project. He predicted it would take 15 years to sequence, while others said 100 years. Who was right? Kurzweil. All because he is right about the exponential growth of information technology.
not everything he says, but a good part of it. I’m not saying i believe we’ll go 100% solar in the next 16 years, but I think much of what he says about exponential technological growth – despite how bizarre it sounds now – will end up being correct. We don’t really acknowledge enough how amazing our society is right now. In less than 2 centuries, we went from little mechanical systems to MacBooks and inch-long drives that hold a terabyte. It took humanity 4 million years to figure out how to grow corn! So although some of what he says is a little sketchy, I think he’s more in touch with reality than most of us
As a planning tool, projections are useful but not definitive. The concept of exponential growth is proven in statistics and can be used for extrapolation. It proves that solar power is going to be a major power source in a distributed way. It will be feasible for all new buildings to have solar power installed and with batteries it will be cheaper than supplying gas or oil for heating. It this price point happens in 10 years or 15 years is moot. The point is that for planning huge power projects, the impact of solar power has to be factored into the equation.
Actually, Kurzweil is not basing his statement on Moore’s Law, as you so deceptively claim, rather it is based on the Law of Accelerating Returns. While Moore’s Law refers solely to transistor size, the LAR states that ANY technology which becomes an information technology advances at an exponential rate. Kurzweil has proven that the LAR has held true for almost all of human history, so there is no reason to believe it will stop or slow.
In other words, it’s bullshit. There’s no such “law” at work that Kurzweil has “proven” to function in the world, nor is technological progress inevitable.
There’s nothing wrong in hoping for things, or to state some rules of thumb about the way things go. But if you want to know where science and technology will really be in the future, you’ll just have to wait for it to arrive.