There are many forecasts of the world out to 2100. There is the IPCC climate change report or UN population projections. The projections assume that all current trends continue for the next 80 years.
The assumptions are that there will not be any major technological or social developments that are big enough to alter global trends.
The IPCC and UN did not predict the rise of China’s economy and later under-estimated the speed of China’s rising economy.
The current global economy is about $90 trillion. It has been growing an average of 3.1 percent annually for the next 12 years. continuing this trend GDP would rise to about $125 trillion by 2030.
To get an idea of how powerfully AI could affect economic growth, let’s assume that the last quarter’s 4.2 percent GDP growth rate in the U.S. was somehow sustained through 2030.
A McKinsey Global modeling of the impact of AI is that it could boost world GDP by 1.2% per year through 2030. This would be about double the effect of the internet and Information technology.
The first wave of robotics had a significant but smaller GDP boosting contribution.
AI could potentially deliver additional economic output of around $13 trillion by 2030, boosting global GDP by about 1.2 percent a year. This would be four times the boost of 0.3% GDP per year from the steam engine.
Nextbigfuture does not expect the growth impact of AI to stop at 2030.
Nextbigfuture also expects more powerful robotics, factory improvements and the emergence of true molecular nanotechnology to further increase the economic growth.
There will be factory or shipyard mass-produced nuclear molten salt reactors and various forms of nuclear fusion.
There is a GP-write project to reduce the cost of engineering genomes by 1000 times.
Combination gene therapy modification of adults could massively longevity, health and intelligence.
Boring Company will enable rapid infrastructure improvements via tunnels. Self-driving cars and trucks could boost the volume of commerce by 20 times with more products moved at higher speeds.
SpaceX could enable one to three-hour travel anywhere on the planet.
Tens of thousands of laser linked satellites will boost the speed of communication.
China showed that it is possible to have 9-12% GDP growth every year for over 30 years. Other countries have been in the 5-8% annual GDP growth levels.
An extra 1% of GDP growth for ten years is an extra 15%.
An extra 2% of GDP growth for ten years is an extra 27%.
An extra 3% of GDP growth for ten years is an extra 33%.
Growing at 10% per year means increasing by 2.59 times instead of 1.36 times every ten years.
It will still take some time for the various technologies to fully emerge and interact.
If the world economy accelerated by an extra 1% each year each decade, so that there was 4% per year GDP growth through the 2020s, 5% per year GDP growth through the 2030s, 6% per year GDP growth through the 2040s, and so on. There could be 10% per year GDP growth for the 2080s. A century of super-technology would mean a world economy that is 30 times larger than the 3% flat scenario.
A world with advanced AGI, super-robotics, molecular nanotechnology and abundant energy via nuclear fusion versus a world without that technology would be like Singapore ($64,000 per capita) versus Cambodia ($1500 per capita) or Vietnam ($2500 per capita).
A world with 3% annual GDP growth would have an economy of about $1000 trillion. This would be most economies with per capita income at least at current US levels and some places with triple current US income levels.
A world that accelerated to 10% annual GDP growth would have an economy of about $28000 trillion. This would be a world of per capita millionaires. How could an economy with AGI, massive space economy, molecular nanotechnology and anti-aging not have a transformed level of wealth?
Most of Asia stayed poor through to 1980. Most of Africa remains poor.
It is possible to screw things up and have a mostly stagnant future.
I think the big technology happens and there will be constant transformation. Rapid development and growth requires constant massive change. It would be a world of SpaceX instead of a world of United Launch Alliance (Lockheed-Boeing).
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
53 thoughts on “Supertechnology Will Define the 21st Century and Not Status Quo Trends”
He was a physicist, he probably had no idea if anything he was theorizing would even be possible. At the time of the Wright Brothers there were articles calling flying machines impossible, yet at the time they new physics for one check out. Also I really have no knowledge of physics beyond computer engineering so anything I say regarding aeronautics may be wrong.
Climate change deals with changes to the climate. Which are caused by average temperature changes. Which are caused by Earth’s energy balance. Greenhouse gases alter the energy dissipation rate (energy dissipates more slowly). Excess energy alters the energy addition rate (more energy to dissipate), which is just another factor in the same equation. The net result is the same.
If you’re collecting solar on Earth, then you’d be correct. Nothing is added. But if you’re collecting in space, you will almost certainly be collecring light that would have otherwise missed Earth. If you then beam that energy to Earth, you would be adding energy to a limited system.
There’s a limit to how fast Earth can dissipate excess energy, even if that limit is high. So at some point you will reach a scale where you’re adding energy more quickly than Earth can dissipate, and you start affecting the climate. As I wrote, my guess is that point is somewhere around a few percent of Kardashev I, which is only 2 orders of magnitude above current global energy production.
We aren’t harvesting enough energy for such effects yet, but see also my reply to Jean Baptiste.
(P.S. Noticed all of your replies start with “It doesn’t matter”. Surely, some things do matter?)
Depending on developments, we may have to start thinking about it this century. Brian talks of reaching ~10% annual growth with the various super-technologies. We might get some of these around mid-century, and just 50 years at 10% annual growth is over 100 times total.
GDP and and energy are closely linked, so energy will have to follow for such growth to continue. We can expect some improvements in efficiency, but even so, by 2100 we may begin to approach the point where we’ll need to consider where we place all that energy.
Luckily, we will likely see some major space development this century, and may begin to offload some of our industry to space (large-scale space solar isn’t even possible without that). So part of that energy production and use will naturally be off-world. But by 2100 we may need to shift to space more deliberately.
Btw, when we get nanotech, we could jump straight to Kardashev I just by building J Storrs Hall’s weather machine. But that would only be Kardashev I according to the level of control, not so much according to energy available for industrial and daily uses.
OK, fine, I’ll edit it for those who can’t figure that out…
I don’t think the approach to AGI is really anything special compared to narrow AI. When we have the computing resources to connect multiple disparate AI’s together, you essentially get AGI. AI is in the beginning stages of penetrating and causing disruption to every major industry. AI is getting more flexible and gaining the ability to connect two or more different cognitive processes together to be able to perform complex tasks.
Just as I don’t think AGI is really all that special, I don’t think conciousness in the way humans understand it is all that special either. If a machine is able to generate and work with complex thoughts, draw from memory, and recognize how it’s actions interact with outside objects, I think you essentially are demonstrating conciousness. AI is in many ways already doing the things that demonstrate conciousness. Their level of conciousness isn’t at the level of humans yet, but I don’t think human conciousness is all that special that robots or AI can’t demonstrate a similar level of conciousness in the future.
Even from the perspective of a decade ago, we’re accomplishing things now that would have been considered practically impossible then. At the same time, there were several technologies that were “only a few years away” that are nowhere close to practical applications. Some of these technologies will require a lot more investment before they become practical and some ideas will die out completely.
I guess one thing you can take away from this is technology is hard to predict and it’s hard to tell what the impact on society will be. “Sure thing” technology investments may be a complete failure while visionaries and innovators like Elon Musk can take impossible ideas and make them a reality. I think it’s a safe bet that the world of 2100 will be a far different place than we imagine.
It was around the mid-to-late 2000’s when I first started following technology trends closely. At that time, a space elevator was considered to be the solution to cheap access to space. Also space related, an SLS type program was proposed and had people excited it would eventually take us to Mars. Practically noone thought it would be possible for a private space company with a reusable rocket to completely transform cheap access to space.
Nanotechnology was the big buzzword of the day. Practical applications that would transform the world were only a few years away. AI was not much above the level of a pseudoscience. If someone said in a decade there would be a huge push for self driving cars by almost every major auto manufacturer, they would have sarcastically replied “where’s my flying car?” A Google duplex level AI with a voice practically indistinguishable from a human would have been considered impossible.
Tokamak fusion reactors were considered society’s best hope for a future of clean cheap energy. Few people saw how rapidly the cost of renewable energy sources such as wind and solar would come down. It should be obvious by now that smaller fusion projects focused on novel techniques or even conventional fission reactors are a better approach than the massive international ITER project.
“Improving” is a strong word for AI. Rodney Brooks is very skeptical that current deep learning tech can get us to AGI (Artificial General Intelligence). So the jury is still out on how much generalization one can get out of “line fitting” math.
But if you worked out the basic mathematics that actual anti-gravity people used in the design of antigrav then I think you would have a much stronger claim.
And if the actual anti-gravity textbooks had you listed as the person who did the basic theory.
And if the anti-gravity researchers keep using your equations and referring to them as the “Sprunk Soda equations”.
That wouldn’t matter, we aren’t harvesting nearly enough energy to bake the Earth (climate change doesn’t count since it deals with greenhouse gases and not energy). Energy is released into space and is dissipated extremely fast, us using solar isn’t going to change anything. The energy already bombards half of the Earth all the time, harvesting it with solar doesn’t add or take it away, it remains 0 net gain.
Even when we make an AGI we wouldn’t be able to understand if it even is one if that makes any sense. An AGI is code that can adapt to its environments which we have already with things such as chatbots of virus’. Point being it doesn’t matter if we have an AGI, the distinction would be very vague anyways and would yield the same results as todays AI.
It doesn’t matter, technology advanced and becomes cheaper. The average minimum wage worker lives like a middle class person did in the 60’s, even the poorest people with houses in the country can afford fridges and other tech. A person with minimum wage in the 60’s couldn’t afford nearly anything.
He thought of it, that doesn’t mean anything. Just because I think of anti-gravity doesn’t mean I would have anything to do with its invention in the next thousand years.
So we still don’t have it
It’s clear that we shouldn’t try to reach Kardashev L1 by baking Earth in microwaves and adding any significant fraction of Earth’s Solar energy input to it.
For growing beyond some fraction of L1, we need to expand into space and distribute that energy into a bigger space.
Or reduce Earth’s insolation (by making Sun shades), but that would be somewhat unsightly and undesirable for many, me thinks. People (except astronomers) can maybe tolerate some new stars in the sky, but not big patches of shadow and night where previously there was sunlight. Not without being forced to, and by that time they will certainly have other options.
This is of course idle speculation today, but it may not be so a few centuries hence.
Yes, Kardashev I is about 10000 times current global energy use. That would be the entire natural energy budget of Earth. The point where added energy begins to alter the climate is probably somewhere around one to a few percent of that. So about one to a few hundred times current global energy use. Two orders of magnitude is within the scope of this article (though it is talking about GDP, but GDP and energy are correlated).
There’s a similar limitation with nuclear. Wind and ground solar don’t have this problem, since they come out of the natural energy budget, but they can’t be practically scaled anywhere near Kardashev I anyway.
Not sure if that holds in a post-labor society. If there is still human labor, it may not pay much. And there may not be much labor available. Yet certain products will remain at a premium. Select real estate is a good example. Maybe some non-essential products or services that have limited supply.
The amount of power required to heat the Earth up directly (as opposed to mucking around with radiation transparency in the atmosphere) is orders of magnitude larger than we are currently dealing with.
I suppose it depends on the meaning of “practical limit”
Say’s law. By definition there will be as much income available to pay for labour as there is labour being paid.
Hmmm… rich and famous professional athletes were a thing in Rome.
Your ability to get a better job does depend on GDP growth. In combination with your own drive and ability of course.
OK, this is probably some regional difference in the meaning of dream and imagine, they are synonyms in my country.
Secondly, to quote the wik
He was most definitely a scientist, and developed most of the basic theory used in rocket design today.
> no practical limit on quantity
There is: beam too much energy down to Earth, and you’ll be causing climate change just by thermodynamics. Any more than that we have to use in space.
There may also be lesser engineering limits such as how many power sats you can stick in a convenient orbit, or where you can place the receivers.
Nonsense. I will address each.
Trump happens. Regime changes happen. Currency reforms happen. Tech reaches the upper-right corner of logistical curve and is discarded. Cycles everywhere, with occasional stochastics.
They’ll find something. What matters to any these ideologues is the restriction of human freedom. Any tech factorizing human beings out of the ecosystem and its resources and therefore, out of their direct control, is their principal enemy.
But it may become increasingly hard to sell as a control mechanism. After all, powering our civilization with solar power is the sum of everything these people have been fawning over for decades.
Maybe they’ll focus on policing and punishing wrong think and wrong tech (e.g. pesticides and meat farming). Fighting much more abstract demons has done wonders for decades for them.
Trend towards faster, and more useful information technologies.
Trend towards more, and more worthless currencies all over the world.
Destruction, reinterpretation, or what ever other lie to take away the rights of American citizens.
Lie de jour, the American president is not in charge of diplomatic relations. The lie is made in an attempt to remove a duly elected president, while claiming to support democracy.
The single most transformative medium term technology is power satellites. There’s pretty much no practical limit on quantity, or delivery, except near the poles if using equatorial orbits. I suppose it’s natures way of saying, “you should use that heat sink”.
Replacement of other energy infrastructure with microwave, or laser receivers raises a question. If there’s negligible human CO2 emission, what will the globalists use to try to control people’s lives? Methane? Mexican quisine could be made illegal! Tacos rule.
These growth rates can’t last (gdp, population, etc) because the natural tendency is flat. We are living an exceptional trend that is like an avalanche, but avalanches don’t last…
More or less agree, but I think the AI solutions will get increasingly flexible and generic as time goes by. Like we have better machine learning now than 10 years ago.
New problems pop up at a limited rate, I think, and within a limited scope. But once we have that AI apps platform, solutions can be developed fairly quickly. So I think anything that can be solved with AI, will be solved. And that threshold of solvable problems will only expand as the AI libraries and hardware get better and programmers get more creative. Even creative tasks are being tackled by AI researchers.
But even if a problem can be solved by AI, there could still be humans working on it too. The artisans, as you correctly propose. And the hobbyists. They may not get paid much, but if they like what they do, there’s reward in that. And if they’re good at it, it can still be useful to others. Maybe we’ll see government subsidized employment programs just to keep people busy.
With these and maybe other arrangements, there will still be room for work even in a post-labor society. It just won’t be necessary for survival. With wide-spread automation, the cost of covering basic needs could be quite low. Mining, manufacturing, agriculture, energy, transportation, retail, even construction could largely be automated. So common physical products could get very cheap.
What I’m less sure about is the economics of luxury goods. How does one pay for them if there aren’t many well-paying jobs that one can do?
If only 1% of global wealth reaches the bottom, but global wealth is a 1000 times bigger, that’d still be a 10 times improvement for them. A x10 improvememt means the poorest countries get to about the level of eastern europe today, and eastern-europe-level countries get to the level of today’s richest countries.
At least some means of production can be duplicated in open source, so the non-oligarchs can gain access. There are already open-source 3D printers and open source robotics projects and plenty of open source software.
And there are several candidates for abundant clean energy:
– Solar and wind etc. There’s about 10000 times as much solar hitting Earth than all the energy we currently use.
– New nuclear fission, which is expected to be much safer than older designs, and could produce less waste or even burn current nuclear waste.
– Space-based solar will become affordable in the coming decades as launch prices are reduced by SpaceX and others, and we develop space mining etc.
– Maybe fusion. Though even if the technicalities are solved, its economics are questionable.
– Worst case, we can even use fossil fuels while they’re still available, as long as we capture and sequester the CO2 and other emissions.
The energy from these sources has to go to market, otherwise they’d be useless even to the oligarchs. They can keep the power plants, but they can’t keep all the energy to themselves.
Re: your last point – we wouldn’t have to go back many years to find pretty much everyone agreeing that no one could make a small fortune playing <Video> games professionally.
Not quite. The extra 2 billion will be mostly Africa and India, and they are not a part of any bright future for a number of domestic reasons. All energy tech, if one does not count forest burning as tech, is not theirs. Containment of those new 2 billions, and their busy parents, is a known darkening factor for others, but it is a simple matter of reasonable domestic governance in the containing countries.
Oligarchs can only control economy, they cannot divert its output into themselves. As such, it is also a simple matter of domestic governance to motivate oligarchs for constructive activities. They can own fusion plants and robots, if those assets are engaged in constructive activities. Someone will own them in any case.
Dreamed of, not imagined. Same as “warp drive” is dreamed of today, not imagined. He was a school teacher, and to imagine a rocket, one must be an engineer, also a scientist if that is the first time.
Yet we have clean cheap energy almost here and now. MSRs – the only thing stopping them is Leftist FUD and that the politically powerful do NOT want a solution to the energy problem that would take a away the “crisis” of climate change.
Your own words about our ability to project into the future should be applied to your ability to look into our own history. Much of what you said is your own bias and not based in fact, scientific or otherwise. Doctor Pat gave you one simple example. It seems to me you are the one blatantly making statements and getting it wrong. Not necessarily all of it.
Do not despair! Modern medicine will bring a cure to your persecution and radical conspiracy theories. It’s just a matter of getting your brain chemistry in balance.
As far as the oligarch class, you are already a member. We live better than 93% of the rest of the world. You’re just too sick to know it.
What does that even mean.
Big rockets were first imagined in 1903 by https://en.wikipedia.org/wiki/Konstantin_Tsiolkovsky
Lets look first at things we can see coming:
Fusion power. Despite the refrain of its always 50 years out (and lately 30 and 20) the reality is that multiple groups are working on it, and it looks doable within 10 years or less. This changes a ton of things. Everything is cheaper to use or manufacture. Many of the designs currently underway woud be able to run planes, ships, and trains. Vastly reducing their cost.
AI. Its improving at a vast rate, a generic ai would be insanely capable of accelerating change and development, but even specific purpose ais such as self driving cars, and medical diagnostics are huge.
Robotics. More general purpose robotics that are reduced in cost due to their ability to be mass manufactured would change tons of things.
Genetic manipulation. crispr and now a even more capable one thats recently came out open up fascinating possibilities in health care, manufacturing, etc etc.
Nano assemblers. Recent work on this topic has seen some improvement in design capabilities, and I think someone with a couple billion could do this in under 30 years. maybe under 10. And this one is huge. Because in addition to assembling things, they could disassemble things. Wreck your car? bring it home, put it in your garage sized assembler….disassemble it, and reassemble it as the newer model.
Now most folks are thinking of these as standalone things. They arent really. Start thinking about the combinations of these.
Without clean and cheap energy there is no next big future.
Any kind of mass manual labor will be toast as soon as a machine can do it for cheap(er). This has been the case since the industrial revolution started.
But then, artisan labor will probably become appreciated just because it’s human made.
Same for technical, creative work. The set of problems to solve seems to always keep ahead of the set of existing solutions for any given problem. The main reason for that are the unforeseen interactions of new technologies and developments, creating whole new problem categories.
This besides the fractured, ultra-specialized nature of AI solutions, means someone with a mind and a will is going to be always necessary, despite the existence of an ever growing set of automated problem solvers.
We are seeing the start of that now, with the ever growing crunch of technically skilled workers, and the push towards ‘inclusion’, which really is about trying to gather all those willing and able to fulfill some necessary role.
I can see lots of products meant to be used, begging for developers, creators, entrepreneurs and really, someone to know and care about them.
I want No. 2 to happen so bad
Without clean and very cheap energy, a very bright future for 10 billion people will not materialise. Even with those advances, a majority of the economic pie could go to a small oligarch class who owns the fusion reactors and the robots.
You’re welcome to live your plebeian life any way you see fit, most people’s outlook in this world is defined by poverty.
Your wage does not depend on GDP growth rates if you make minimum wage.
Technically, we don’t need AGI for a largely post-labor situation. A generic robotic platform with a decent OS and good libraries will do. It needs to have good sensors and good dexterity and be easy enough to program. Given the current state of leading-edge prosthetics, I think we’re not too far from having good enough hardware. The OS and libraries may be more difficult, but once we have that, there can be an explosion of task-specific narrow AI apps, like we’ve seen with smartphone apps. The more advanced the hardware and software, the fewer jobs will be left for people to do.
I do expect people would still work even in a post-labor economy, but it would be mostly for self-satisfaction and maybe to pay for luxuries. Though I’m not sure how the economy of that would work.
I’m also not sure full-blown APM is needed for something like the above post-consumerism scenario. Advances in additive manufacturing, modular robotics, and miniaturization may be enough for a limited version of that. Something like mm-scale claytronics (modular robots) being able to form some useful products, advanced additive manufacturing able to make some others, and you still have to buy the rest from traditional manufacturing sources.
But 80 years is a long time. I expect we’ll get full APM by the end of this century, maybe by the middle of it. Less sure about AGI, but maybe same. A generic robotic platform with narrow AI apps probably sooner.
So nice to add annual percentages and see how accumulated growth explodes. We are constantly adding new technology, the internet in the 90, cellular communication in the 2000s, yet global growth rates stay constant. There are absorption limits. Developing countries can grow faster than mature ones because they are adopting technologies that have been tried elsewhere.
Global Annual GDP growth:
Well, yes. I pretty much as take it as a given that every major issue we face right now will be OBE (overcome by events) by mid-century, thereabouts.
Look to the past and you will find many disasters that never came about.
Not to panic, we will have plenty of new things to panic about. The Billy Joel song, “We didn’t start the fire” seems especially apt here.
Also, while I don’t believe it is especially prophetic, reading the book Accelerando by Charles Stross (which tells the story of one family’s passage through what some might refer to as The Singularity) might very well convey just how radically things may change, and the incredible pace at which they do.
It’s fun to play around with. Even the non-dystopian possibilities are endless.
While AI has certainly advanced a lot, AGI is still nowhere to be seen and it doesn’t seem likely to arrive soon. It seems as if AI is rather going towards a myriad of fractured applications, with different problem solving approaches.
Which for me is totally fine, being human and appreciating having a job and living.
Nanotech is also advancing, but its time to shine hasn’t yet fully arrived. Maybe it’s something else to look for on this century?
Trends do not last 80 years. 80 years ago WW2 barely started, no one could imagine a big rocket, space rockets were fantasy, there was no nuclear power or weapons, five vacuum tubes in a box were high technology, and the best computer was built with mechanical logic. No one is capable of projecting at 80 years into future without blatantly making stuff up and getting it all wrong. The future is unpredictable – it is just a scientific and historical fact.
This is true, esp when you realize that “the world” includes G. K. O’Neill settlements, on Amazon.
Note that there is at least one study that shows China did not have the growth rates usually believed. And, much like Britain in the early industrial revolution or the United States, post WWII -Chinese economic growth was the product of a specific time and set of circumstances. The three of them taken together show that high sustained medium term growth rates can occur in specific countries, but do not show that such growth rates can or will occur nearly universally.
Also if it was true that trends continue as projected, based on historical trend projections I’d have my flying car by now and be planning on vacationing at Hamilton’s Hazard in L-City on the Moon.
Just AGI alone (coupled with robotics hardware advances) can bring about a post-labor economy.
Just APM alone (Atomically Precise Manufacturing, i.e. full nanotech) can allow most physical products to be formed on demand (if we can make utility fog) or almost on demand (with nanotech-based 3D printers, if utility fog turns out impossible or impractical) from digital files. The former would have a marginal cost of effectively zero: just the energy cost to rearrange the foglets.
The latter would be close to zero: energy + materials. The materials can be mostly CHON, which can be obtained from just air and water. Maybe add cheap organics like wood scrap or food leftovers for a more concentrated carbon source. Or the stuff that goes in your toilet – less sewage waste that way. Fully automated collection, disinfection, and recycling, so you don’t need to handle it. So energy + air + toilet stuff for the 3D printer case.
Either of these nanotech scenarios could make most traditional manufacturing obsolete, and turn consumerism on its head. Rather than going to a shop or mall or ordering stuff online, you’d only buy a utility fog package or 3D printer once, then download the design for whatever physical goods you need. Each of these could cost next to zero. Post-consumerism.
Combine these, and the resulting economy would be so drastically different, that comprassion with today would be next to meaningless. And that’s even without any of the other things than Brian talks about.
Comments are closed.