Sorry, I'm back. I've been keeping myself kinda busy over the last two weeks. On my travels I met what you could consider to be a pretty bright programmer: he writes spider programmes. Now if you were silly enough to want to sit in the first few rows of a concert from some mediocre but popular pop star, you would probably want to be cursing him: for his boss and his spider programme would already have the tickets. He works for an entrepreneur in a nameless but extremely large country, who buys up all the tickets for 250 dollars and re-sells them at around a thousand a go. He told me that at first this work was easy, but recently things have gotten more difficult. The concert organisers have tried to overcome the practice by having an image inserted to which you have to manually type some given response. Problem solved you might think. Well no: this is where ingenuity and globalisation come in to guarantee that 'real' entrepreneurship will not be thwarted.
His boss responded creatively: he contracted a hall with 200 workers in India. These workers spend their day typing the image responses manually into a data base. Currently they have entered something like 500,000 images. (It also occurs me that systematic spam must do something like this: the bacteria-antibiotic effect). The recounting of this story lead my Argentina blogging friend Marcelo to make the following highly perceptive observation:
Now I think he really has a point here. The internet skeptics are so busy being skeptical that they don't notice when the roof is falling in around their heads.The image of the wharehouse of people defeating the turing-test safeguards is extremely interesting. At the risk of sounding callous, I think that an interesting way of conceptualizing what's happening in India and China is that Moore's law is applying to humans: the capacity of a person you can rent for $1 is increasing fast, thanks to a bigger pool of people and better technology to teach and connect them. Of course the pool of people is finite, and eventually you start getting higher wages, but the principle is the same - and if stuff like MIT's Open Courseware works well, the trend might well continue.
On my website deflation page I identify three factors which might be contributing to a global deflationary environment: OECD ageing, surplus labour in China (and now, increasingly, of course, India), and the falling price of information. Now I have never really been to clear where to go with this third one, it was more a case of reading Kurzweil and extrapolating what to me was the obvious. Now Marcelo has come along and put it very succinctly: Moore's law as it applies to humans. And like the other version of the law, the only remaining question is how long can this run till we hit specific physical limits. I think Kurzweil's answer would be: farther than you imagine.
No comments:
Post a Comment