the skeptic’s guide to the singularity
Whenever I write about Ray Kurzweil and the imminent paradigm shifts in technology and human evolution he predicts will happen in the next few decades, I always get a stream of very critical feedback calling me way too pessimistic, or arguing that new and better technology will always be possible as long as there are engineers working on it. The response from the Singularity Institute has varied from critical, to extremely congenial with the important point that we shouldn’t shoot down the promise of highly advanced computer technology but do our best to explore it. And you know what? I agree with all the points brought up by the critical readers and by the members of the Institute. However, I have a number of issues with Ray’s fantastic world of tomorrow and want to provide a cold, hard, skeptical look into the science and technology behind making it happen in reality.
Believe it or not, I actually want to work on some of the projects brining some of the sci-fi style machinery Ray constantly invokes to life. But my problem with the Singularity blueprint for the future is a rush towards seeing the progress happen and escaping mortality since this is Kurzweil’s primary goal. Transhumanists want to be immortal and they try to find any reason why this should be possible. The end result gives them an alternative worldview in which technological advancement somehow becomes part of human evolution and the idea of a tool that could be used to radically extend our lifespans and grant us the ability to branch out beyond the Earth as some sort of imminent evolutionary step by which we transcend nature and take evolution into our very own hands. But that’s not how evolution works. Rather than take the process into our hands, we’re still going to be ruled by it no matter what we do.
Instead of planning to get away from evolution and perfect nature, we should be trying to learn from the bizarre and nifty solutions that evolved to counter harsh environments and restrictions placed on us by chemistry and physics. For a very long time, we’re not going to be giving birth to cyborgs. Rather, we’ll be turning people into them as they grow and age. We’ll need to keep our current organic brains to allow the random complexity and subtle interactions of neurons to shape our personalities. We’ll need to limit tinkering with our genomes to a minimum so we don’t rely on rather flimsy gene therapy and let evolution do what it does. And very importantly, the engineers who’ll be working on these solutions need to plunge deep into the realm of science. When we start working on nanobots to help blast away cancerous tumors and perform surgeries without scalpels, we’ll need to spend years studying biology and medicine with qualified scientists as our guides. For those of us in the computer science realm, we’ll need to inject as much science into our work as we do engineering to truly provide Singularity-style solutions.
I simply can’t understate the importance of realizing that every transhumanist desire is very firmly based in our evolutionary drive to survive and ensure our species continues. Every one of the changes we want to develop so we can walk on other planets, become as resistant to radiation as the heartiest bacterium, and maybe put death on a very long hold, would just be a response to natural selection. We’re never going to transcend every rule imposed on us by the laws of nature. We just need to make the best of them and with enough effort, time and money, we can really build amazing things. But we shouldn’t invent ridiculous dogmas and pretty charts devoid of practical meaning. It’s probably hard to tell from my critical posts, but I’m very passionate about this stuff and could easily spend hours building requirements for some of these technologies, discussing all the pros and cons of mind uploading vs. extensive programs to change humans into cyborgs as well as the bugs and technical challenges we’d need to address. I’m even continuing my education for a chance to do tangible work in this direction and apply what I’ve learned about science to the realm of speculative technology.
So ladies and gentlemen, I don’t simply snarl at the Singularity or dismiss the work slowly being done by the idea’s proponents. But just like scientists want to critique and debate ideas to find the right answer, designers and engineers want to weed out potential dead ends and clear up misconceptions. This isn’t about optimism vs. pessimism or practical vs. pie-in-the-sky ideas. It’s about separating real projects from wishful or suspect thinking and creating real blueprints, real requirements and getting down to work. The longer we talk about all sorts of pseudoscientific transhumanist mythologies, the longer it will take to bring the products their want to market and accomplish the lofty goals from which humanity will benefit.
[ illustration from a matte painting by Christian Hecker ]
We’ll need to limit tinkering with our genomes to a minimum so we don’t rely on rather flimsy gene therapy …
Whatchoo mean “we”, white man? So long as there is a dollar/euro/yuan/Federation Credit to be made, expect genetic tinkering to proceed much faster than ethical frameworks and regulatory structures and comprehension of social/ecological impacts.
Please note, this prediction is based on both deeply-ingrained cynicism and extrapolation from existing data points (yes, I’m looking at you, Monsanto!).
If we just ignore Kurzweil’s hand-waving, aren’t we just left mostly with real science and research targets?
Keep in mind that “Singularity” refers to three distinct concepts. People are very familiar with the accelerating change version, which is partially just a repackaging of transhumanism in general, but not so familiar with the other schools.
There is some pseudoscientific mythology in transhumanism, but also common sense. For instance, see “Transhumanism as Simplified Humanism”. Many humanists seem to be laboring under the delusion that there is are magical “hubris ceilings” that we will never break through. One of those is the delusion that human-equivalent artificial intelligence will take many hundreds of years. Such a position doesn’t really make sense unless one views the mind as magical rather than just what the brain does.
“Such a position doesn’t really make sense unless one views the mind as magical rather than just what the brain does.”
You don’t need a spiritual, metaphysical view of intelligence or consciousness to think that AI isn’t going to be a walk in the park, take many decades to develop, and will probably never reach a level comparable to that of a human. I really need to do a full post about this but right now, I can summarize the reasons thusly…
1. Human brains are complex and evolved over long periods of time. Unless we know how intelligence works on a molecular level and can guide a computer step by step, we’re not going to have true AI, just a facsimile that replicates aspects of what we call intelligence. It takes a very long time to truly figure out a system as big and messy as our minds.
2. We would need to invest a lot of time and effort in virtual stimulation so the AI has the motivation to do anything at all.
3. It’s incredibly impractical to set up full blown AI omni-app. The project would have very little use outside of the lab and the ideas of using clever AI to give us an edge in making new generations of technology would rely on the machines analyzing the potentially faulty knowledge we collected. The result? A system with seemingly no real purpose we’re trying to use to sift through data being constantly updated. We’d get a very fancy search engine at best.
…we’re not going to be giving birth to cyborgs. Rather, we’ll be turning people into them as they grow and age…
I can personally vouch for that statement as entirely credible! If it wasn’t for modern cardiac surgery techniques (replacement parts both mechanical and natural) that were formed over a period of decades, I wouldn’t be here to type this.
So long as there is a dollar/euro/yuan/Federation Credit to be made, expect genetic tinkering to proceed much faster than ethical frameworks and regulatory structures and comprehension of social/ecological impacts.
I agree with Mr. Butler on this. Ever since the Gilded Age, technology belongs to highest bidder, no matter what regulations might be in place. Politicians and judges can easily be paid-off to overlook any rule/regulation. If there’s one actually in place to begin with that is.
Politicians and judges can easily be paid-off to overlook any rule/regulation.
That’s absolutely true. But, the statement that we should limit our genetic tinkering is not a high brow principle. It’s just sound science. Gene therapy is one of those ideas that sounds great on paper, but in reality, it’s a big, complicated bag of worms that doesn’t have the kinds of benefits we hear about in pop sci articles.
It’s just sound science.
Very true. But greed will trump it every time.
Maybe Kurzweil’s super-evolutionary Singularity will raise us above such things?
Like you Greg, I’m not holding my breath on that occurring!
“imminent evolutionary step by which we transcend nature and take evolution into our very own hands. But that’s not how evolution works. Rather than take the process into our hands, we’re still going to be ruled by it no matter what we do.”
Haven’t we been doing just that since the times of the Agricultural Revolution? Also, the process of sexual selection could be argued as a form of taking the process of evolution into a species’ own claws/mouths.
“every transhumanist desire is very firmly based in our evolutionary drive to survive and ensure our species continues.”
I cringed a little when I read that line. Individual members might strongly desire the continuation of our species, but I I haven’t seen any evidence that this drive is ingrained via evolution.
These are minor complaints, especially since the rest of your article is about how we can’t transcend nature, and that using it to our advantage is the only way forward — a claim with which I fully agree. But isn’t that supposed to be what Transhumanism is? At first I thought your arguments against Kurzweil were something of a straw man, though I’ve never seen problems with the body of your arguments. Perhaps I’ve been ridiculously blind to the woo woo Transhumanists?