The Final Industrial Revolution Tuesday, Apr 29 2008
futurism 4:36 am
The Industrial Revolution is essentially when the modern world began. For the first time, the production of food and many basic goods was partially automated. The impact was huge: whereas most of history up to that point had been a zero-sum exercise in fighting over a fixed pie, the Industrial Revolution increased the size of the pie itself many times over. This laid the groundwork for positive sum thinking, the notion that we can all have better lives if we just cooperate instead of trying to edge out the next guy.
Today, we have the luxury of pointing out some of the downsides of industrialization, but if we could experience pre-industrial life firsthand (or observe it by visiting those few areas of the planet untouched by industrialization), we would recognize what we are blessed with. Usually we reserve our gratitude for other humans, but the machines that drive industrial civilization deserve our thanks as well. Without them, there would be no mass-produced clothes, or toilets, or plumbing, or medicine, or books, or computers, or pretty much anything except for what we or people we hire could build on a manual, one-by-one basis.
Many of the “superlative” visions of transhumanists, and some other futurists and tech-savvy folk, introduce the notion of a world where the Industrial Revolution is completed, and literally any production we desire can be automated. In the words of the RepRap project, this would give us “wealth without money”. To me, and many others, this is a question of when, not if. I see modern skepticism over the possibility of self-replicating machines to be analogous to pre-Industrial skepticism over the notion of a huge array of products without excessive labor input.
Because I’ve been thinking about self-replicating machines for over a decade now, I’ve already spent quite a bit of time considering many of the positive and exciting implications of the technology. Palaces built by automated legions, cities constructed in weeks or less, solar panel arrays hundreds of square miles in extent covering the desert floor. Call it superabundance, the natural consequence of having machines that self-replicate from raw materials, are powered by the sun, and do all the work for you. It’s the Final Industrial Revolution, because after we develop reprogrammable self-replicating machines, there will be no production left to automate.
I’ve thought about the feasibility and potential of self-replicating machines enough that I’m desensitized to many of the benefits. Maybe I was introduced to the concept young enough in life that I take it for granted. Instead of speculating endlessly about the possible benefits, lately I’ve been very concerned about the risks. My thesis is this: if we can survive the next 50 years or so, we’re home free. I don’t see actively pursuing self-replicating machines as morally necessary as ameliorating the risks, because the eventual invention and deployment of these machines is inevitable. Someone will definitely do it, probably sooner rather than later.
So I consider the risks. Unfortunately, this puts me up as a target for criticism, along with people like Martin Rees and Stephen Hawking. For instance, one “expert in apocalyptic movements and neocatastrophism” called Hawking’s warnings of catastrophic global disaster “regrettable hype”. Honestly, what other intelligent species could build enough bombs to kill everyone on the planet, then go around saying that we have nothing to worry about?
People are terrified of discussing catastrophic risk for a number of reasons, but I think the two primary ones are negative affect and a bias towards optimism. They sound like the same thing, but they’re actually slightly different. Negative affect is the uncomfortable feeling many people get when discussing something they don’t like, such as the annihilation of the human species. The solution is to turn the conversation to a jokey tone as quickly and desperately as possible. Sure, I guess heavy topics aren’t for everyone, but I wish that people could discuss the risk without flinching away at least 1% of the time, but I’m not seeing it. The second reason is a positive bias towards optimism. Humans have a well documented psychological propensity towards unrealistic optimism. Optimism is a badge you can hold up to a community to say, “I’m psychologically normal, and motivated to join in shared projects!” Well, great, but when we’re in a unique historical era where we have weapons that can kill off the human race but lack the wisdom to forgo their use, this attitude can be downright suicidal.
The Final Industrial Revolution will be happening in our lifetimes. How can we steer it down a beneficial path? “Stand back and let it happen” is a recipe for failure.

April 29th, 2008 at 10:02 am
I prefer to call it the _Anti-Industrial Revolution_, because it will eliminate of most of the things we associate with the Industrial Revolution: factories, mass-production, mass-consumption, commercial finance, economies of scale, capitalism…
April 29th, 2008 at 12:50 pm
What I see in the development of nanofactories could turn out to be the final war between capitalism and communism. Capitalism relies on industries, supply and demand, and consumers. When, however, the consumers become the producers, the system can no longer function as it is. Communism, or a kind of neo-communism, or something entirely different designed by enhanced intellects would have to take over in the place of capitalism. The question is whether the wealthy, free governments, and the free market system can adapt to the final industrial revolution. One scenario, the wealthy emigrate to space and leave the rest of us Earthlings to figure it all out. Which wouldn’t be so bad.
April 29th, 2008 at 3:19 pm
If all goes according to the optimist’s plans then future is looking peachy, but I agree that optimism in the face of this essentially negative reality(our ape brains handling self-replicating machines) is going to lead to death for everyone. We do need to face head-on the coming dangers if we hope to survive. I’m not doing anything proactive to keep these things from happening myself but I congratulate you Michael for being so brave and proactive about these challenges. Especially when talking about this subject matter is so stigmatized.
“One scenario, the wealthy emigrate to space and leave the rest of us Earthlings to figure it all out.” remember when the simpsons did that in “Life’s a Glitch, Then You Die”. Albeit, it was an apocalyptic scenerio in that episode.
April 29th, 2008 at 8:19 pm
The major difference between the Industrial Revolution and the coming technological revolution is that we won’t have decades to acclimate to this one. We might not even have a decade.
April 30th, 2008 at 12:49 pm
Once machines start reproducing themselves then the forces of evolution take over, and evolution is a notoriously difficult process to micromanage. However, just as in the natural world reproduction will require materials and time. Our best hope if we wish to regulate the pace of machine evolution may be to carefully ration the availability of resources needed for their reproduction and to design them initially such that they are dependent upon the sorts of resources which we can reasonably control.
April 30th, 2008 at 3:27 pm
Bob… you… actually answered my question. It’s rare to see that.
Why do you associate self-replication with all of the “forces of evolution”? Biological evolution is a highly complex, specific thing. Darwinian population genetics. When you say the term “evolution”, I think of that.
Surely machine replication won’t be the same thing as population genetics. I should think that intelligent humans, such as ourselves (bwuahahaha) should be capable of programming machines that only self-replicate along the lines we desire. I.e., if a cosmic ray strikes the robot’s programming, its effects will merely be repaired by software, or, if the code is irreversibly irradiated, then the robot merely shut down.
A superintelligence could pay very close attention to a lot of things at once, even quadrillions of robots.
April 30th, 2008 at 7:22 pm
What Michael said: No Evolution for Corporations or Nanodevices
I don’t think accidental out-of-control replicators are a likely danger; designing a replicator that can work in a variety of natural environments is probably very hard, and so obviously dangerous that you’d have to want to destroy the world to do it.