Friday, February 29, 2008

Would a technological singularity really happen?

I am not sure how many of you are familiar with the concept of "technological singularity". I wasn't as well, till last Thursday!

Singularity is a state (from Kurzweil's The Singularity is Near) when human intelligence transcends the barriers of the laws we govern our physical and virtual realities with. It is a point which is predicted to arrive in this century where there won't be apparent differences between man and machines any more; because we would be capable of designing ultra-intelligent machines. Our intelligence coupled with the machines, humans would take technological strides not only on Earth, but the entire universe. As Kurzweil's book describes it,

"Now, in The Singularity Is Near, he examines the next step in this inexorable evolutionary process: the union of human and machine, in which the knowledge and skills embedded in our brains will be combined with the vastly greater capacity, speed, and knowledge-sharing ability of our own creations."

And ironically this first ultra-intelligent machine would probabaly be the last invention would ever make; as the world would then be taken over a more superior species: the species Man would make himself through technological evolution (compare, biological evolution which yields homo sapiens).

But I am skeptical about the advent of this kind of a technological singularity at least this century. Here are a few reasons for that:
  1. I agree we have been increasing our technological capabilities at a accelerating rate: a form comparable to Moore's law. But depends how much extensible it is: is there a threshold to it? We still don't know the future of the exponential curve of technological progress.
  2. Now the comes the capabilities of Artificial Intelligence in playing a significant role in reaching the Singularity. Look up the kind of AI paper that used to be written in the 90's. Most of the visionary or theoretical papers' goal was how to design that ultimate intelligent machine: which can perceive the world as we do through our five senses, reason, learn and take intelligent decisions. And now see the AI research today: most of the research has been segmented; people work on building an intelligent system which can either analyze human speech, or computer vision, or a robot navigating a rough terrain and taking optimal decisions for movement. The concept of intelligence is now distributed; we are trying to build a host of intelligent machines which render a focused task intelligently. However our knowledge or capabilities of how to integrate all those intelligent machines into one single ultra-intelligent one is still very limited unfortunately. The future of this kind of integration research is also pretty bleak!
  3. The achievement of morality. The book says, we might be able to get over the morality that characterizes us today. But the question is, one the way, aren't we giving rise to more deadly diseases? Isn't a significant part of the research diverted to that?
  4. The only hope I see that can render technological singularity is nano-technology or bio-informatics. The hope to see an ultra-intelligent machine is predicated by the fact that genetically it has to be coherent with human brain. The question still remains: can we with our limited brains, build a machine which is smarter?
Only future can tell us...

4 comments:

sodamncool said...

The day this happens...I want to be Morpheus :D ...Being the 'one' is too much responsibility :) ...However, right now, I do feel like a robot though..doing the same things over and over again everyday...monotonous galore...At Singularity point, I wonder if I will be one of them...

Munmun said...

@ sodamncool
Ha! Yeah, actually they say, at post-Singularity, the differences between man and machines would become so transparent that hardly we would be able to perceive it! So keep your fingers crossed! They predict Singularity at 2045: not too long! lol!

scritic said...

Yes, Ray Kurzweil is a techno-utopian and I don't buy much of his predictions either.

But the "singularity" concept was actually first thought of by Vernor Vinge - who's a science fiction writer and it figures prominently in his novels and stories. I believe the best one is "Marooned in Realtime".

Anyway, here's his essay on it:
http://mindstalk.net/vinge/vinge-sing.html

He wrote another one recently called "what if the singularity doesn't happen" - check it out here: http://www.kurzweilai.net/meme/frame.html?main=/articles/art0696.html

Munmun said...

@ scritic
Great! Now that I am reading Kurzweil, I should check out these other stuff as well.
On a separate note, Issac Asimov also did hint at post Singularity in one of his fictional short stories, The Last Question.