This is not the blog you're looking for

I have moved, and you can find new entries, comments etc. at www.sandrewswann.com come over and check it out.



Wednesday, October 31, 2007

Singularity Happens

L. E. Modesitt, Jr. wrote a long post on his blog on why he believes the singularity won't happen. [via SF Signal]

According to Modesitt, it won't happen because such visions are based on technology, not on humanity and they're based on a western European/North American cultural chauvinism.

He goes on to explain:


One of the simplest rules involved in implementing technology is that the speed and breadth of such implementation is inversely proportional to the cost and capital required to implement that technology. That's why we don't have personal helicopters, technically feasible as they are. It's also why, like it or not, there's no supersonic aircraft follow-on to the Concorde. It's also why iPods and cellphones are ubiquitous, as well as why there are many places in the third world where cellphones are usable, but where landlines are limited or non-existent.


All in all I think he makes a well-reasoned and cogent argument that completely misses the point. The point of the singularity is the premise, which I think is valid, that it is possible that a technology can arrive that completely overturns the basic assumptions we use to model the future. AI and nanotech are the oft-used sfnal examples, but history is already filled with basic advances that remapped the entire world to fit them: agriculture, sewage treatment, the printing press, anesthesia, automobiles, air-travel, television, the internet, cell phones.

But my main problem with Modesitt's argument is that it is primarily an economic one, based on the assumption that the basic economic rules are somehow set in stone and aren't manipulated by technological change. That's only true if you're very broad in defining your terms. A product's value is less and less defined by the cost of the materials and labor required to build it, more and more the impetus to distribute technology is to get the end user to buy into an associated service (psst, wanna free cellphone, how's about an inkjet printer, brand new DVR, just sign this contract) and as fabrication becomes more and more efficient, "things" become more like intellectual property where the cost has little to do with the physical object, counterfeits become ubiquitous, and theft starts meaning some basement entrepreneur is making something that looks too much like what you're selling. The labor theory of value breaks down in a replicator economy. Even his points about energy becoming more expensive is one good fusion reactor away from being moot.

Like I said, IMO his argument is basically why the Singularity won't happen. . . right now.

4 comments:

aeros51 said...

With the cheap thermonuclear energy, I'd add self-maintaining machines that do the vast amount of physical labor (thus cheap energy with the little to no human intervention robot produces a "fabrication with abandon" type economy). Nanotech? Maybe, but I am not yet convinced that self-replicating nanomachines will produce the singularity. Like genetic "engineering" we are still a far cry from being able to truely engineering living machines. My guess is that nanotech will be the same way, that it will exist a long time before someone revolutionary figures how how to turn that technology into something to trigger a singularity.

But I've been wrong before.

S Andrew Swann said...

Actually, we don need no stinking nanites. . .

We're edging closer to a replicator economy without them. Rapid prototyping has been around since the 90's and if you think print-on-demand is going to change the publishing industry, think of iPod on demand (and I mean the whole damn iPod, songs included). What happens when I can download furniture?

aeros51 said...

Another person that thinks that nanotechnology may not be the way to the singularity. That's heresy in some circles (but I'm a heretic in those same circles).

Rapid prototyping is neat and has lots of uses (I have used it quite often myself), but it is nowhere near the level required to even think of a singularity. First, the good quality (high tolerance) stuff is made out of plastics which has terrible material properties. They have improved, but they are still far off from a machined metal part. The metallic rapid prototyping is an order of magnitude worse as far as tolerances go and usually require handworking or machining to get a workable part.

Second issue with rapid prototyping: it takes a lot of energy to produce a part, which is fine if you are producing only one or a few parts, but it needs to get really fast and power consumption needs to come down (or power needs to become really, really cheap), especially to compete with casted, forged, or even machined metals.

So, its a step in the right direction, but they need to come up with a whole set of different technologies than those listed to even think about using rapid prototyping widely.

As for the possibility of a singularity itself, here is a "what if": Can the nature of humanity even cope with a singularity. At some point, will human nature itself apply some sort of friction to technological advancement keeping a true singularity from becoming reality.

BTW fun topic.

S Andrew Swann said...

Ah, the whole human nature thing.

That's another potential deal-breaker as far as the on-rushing singularity goes. What happens when technological change alters "human nature" and human physiology?

And I'm not talking cyborgs. We're already morphing human beings beyond their design specs just with changes in food intake and something as simple as shoes and clothing. Throw in pharmacology, and who knows what can happen? We're one or two neurochemicals away from removing the biological need for sleep. Think about that for a little while. What happens when everyone has 33% more time on their hands?