Chicken ITle, robots, and a chessboard

Today our new blog character returns. Chicken ITle draws our attention to those who announce catastrophic consequences for IT if we don't all rush off and do something. Usually these sky-falling folk are either analysts or vendors, because both profit from fear, uncertainty and doubt (FUD). But sometimes they are book authors, pumping a book. In this case it is Erik Brynjolfsson and Andrew McAfee from the MIT Center for Digital Business, and authors of Race Against The Machine, a book which apparently tells us "what the rise of the robots is going to mean to all of our lives.”

Martin Thompson at ITSM Review talks about a podcast with the authors. (You are reading a post about a tweet about a post about a podcast about a book, so don't ask me what the book is really about: I have seen enough of McAfee in the past, and I've no time to read Chicken Little-ish books). As Martin tells it

The MIT professors claim that job creation, which used to peg equally with productivity, has ‘decoupled’. New job creation grew on a par with productivity since the second world war until around the year 2000.
Since then productivity has grown whilst new job creation and income growth has stalled. The professors argue one of the main drivers of this ‘decoupling’ is technology.

We had a productivity gain when we went from hunter gatherers to agriculturalists. We had more big steps with mechanised farming, urbanisation, spinning mills, production lines, and off-shoring. All gave us quantum leaps in productivity with associated social disruption, and lots of noise about how the spinning jennies or combine harvesters would take everybody's jobs. But there are now 8000 times as many of us as when we started and we all live richer, fuller, longer, better lives (quiet, hippy). That's productivity gains right there. We've had enormous increases in productivity right across our history. Technology is transforming our knowledge and communication and lifestyle, but I don't see our productivity changing at a startling rate that is any orders of magnitude different than the past, certainly not in the last decade. It was Brynjolfsson who gave us the productivity paradox of missing technology productivity in the first place, though to be fair he put it down to a measurement problem and found some of the missing productivity later, so I guess he knows a lot more than me, being a professor and all.

Machines took away the job of scything a field and threshing the grain. Machines took away the job of spinning at a wheel all night. Machines took away welding and grinding inside a new washing machine bowl (Seen it. They had to hire a "simple" guy to do it. He was a grinning loon by the time I met him.) Now machines will take away the jobs of sewer cleaners and miners and filing clerks and dog washers and toe-nail cutters. Roll on robots.

As I said in an earlier post about robots (and about the New IT Age where we get these hippy-dippy irrational pseudo-scientific ideas circulating) there will always be jobs for people. Those jobs change just as they did in the transition to agricultural then industrial and now service economies. Only those who don't grasp basic economics will fail to see that there is always a value exchange between people, whatever form it may take. Hey, we could try slowing and resisting the implementation of new technologies. After all, that has worked out so well every time we have done it in the past.

Job creation and income growth have stalled since 2000 eh? Since the authors are professors of economics, I'm surprised they haven't noticed a few other reasons for slowed jobs and income in this decade besides a technology run-away. How about the collapse of nations, fictional money, and the redirection of trillions into the pockets of the mega-rich?

Ah, we're told, but technology growth is so exponential now. The ITSM Review post goes on to talk about Moore's Law and the analogy of wheat on a chessboard. I've talked before about the fallacy of exponential extrapolation in the real world. The wheat-on-chessboard is the perfect illustration of this. One grain on the first square, two on the second, four on the third, eight on the fourth... there is an amazing amount of wheat on the last square. Except there isn't. Ray Kurzweil, who I've talked about before, likes to refer to "the second half of the chessboard". What Ray fails to mention is that long before you get to the second half of the chessboard you have wheat pissing everywhere, pouring off the sides of the board, making it impossible to count let alone pile it. The pile would be bigger than Mt Everest - you couldn't find the chessboard. Even if you could, there isn't enough rice on Earth to do this. Every simple extrapolation hits physical constraints that slow and eventually cap an increase. Welcome to the real world. Kurzweil and Brynjolfsson and McAfee and all the other abusers of practical science can indulge in all the "oh wow man" intellectual masturbation they want, the real world just doesn't work like that. The graph doesn't rocket off the top, and the sky doesn't fall.

From comments below:

It's true that I inferred the content of the book and podcast, sorry. My post is really about the message that [Martin] highlighted on ITSM Review: that job growth is stagnating as technology accelerates away at an exponential rate, and the inference from that is that "the droids are taking our jobs", that they are eating the gains we should have seen in jobs as technology supposedly drove productivity to dizzy new heights, which incidentally is PRECISELY the message of the first half of McAfee's TED talk, which he doesn't rescind in the second half. (In fact he utterly fails to link the first and second halves of his talk. He does nothing to show how the amazing new future for the third world will address the jobs gap he predicted).

Moore's Law is indeed rocking along with no constraints in sight. What will constrain us is the applications and impacts of Moore's Law. Just because processing power accelerates doesn't mean the REAL world can change as fast, which is what Slow IT is about. Processing power is indeed going to leave us behind, but I don't expect the practical implications of that to run away at the same pace.

I too am optimistic about the future, I thought that was the point of my post. I am optimistic that robots will relieve drudgery, that new jobs will spring up to replace the old, and that the rates of change will not be exponential.

More here on robots eating your lunch
See also the safety razor singularity

Syndicate content