Monday, January 9, 2012

When will the future finally get here?

It's interesting how insanely wrong predictions of the future tend to be.  Even people who dream about the future all the time have terrible track records.  It's not that they predict too aggressively, or not aggressively enough.  They just focus on the wrong things.

One of my favorite examples of this is Robert Heinlein, widely considered to be one of the Godfathers of science fiction, along with Isaac Asimov and Arthur C. Clarke.  Heinlein dreamed about the future all the time, but consistently got two things wrong:
·         He vastly overestimated progress in physical science and engineering, especially regarding space travel.
·         He vastly underestimated progress in computers.

The implications of this can be comical.  In Starman Jones, he envisions a world in which we have  technology that can beat the speed of light by manipulating the fabric of space, but computers which are incapable of performing simple mathematical calculations, so astrogators spend their time on flights looking up figures in printed tables and keying them into the navigation systems by hand.

Even when Heinlein speculates specifically about computers, he gets it all wrong.  In The Moon Is a Harsh Mistress, he envisions a supercomputer so powerful that it wakes up and becomes artificially intelligent.  It had so much computational ability that it can calculate the odds of the Lunar colonists launching a successful revolution.  And yet this massive supercomputer, capable of self-programming and orders of magnitude more powerful than anything we have today, is just about maxed about by doing a simple video rendering of a human torso.

Part of the challenge in predicting the future is that technology doesn't improve at regular intervals.  In a given area, it might barely move for years or centuries, then explode forward in the blink of an eye.  Five years ago, electronic books were a joke.  Companies had been noodling with the concept for decades, but nothing they came up with was any good, and it seemed like it might be decades before anything might catch on.

Then the Kindle appeared, and the iPad, and the Nook, and numerous smaller competitors.  Now the question isn't how fast ebooks will grow, it's how long and in what form paper books will survive.  (I think they will survive for the long term, but at a fraction of their previous prominence).

So we've got ebooks.  Are we in the future yet?

Overall, I'd say no.  We're almost at 2015, and the world still seems a far cry from the vision presented in the movie Back to the Future 2.  It got a few parts right, like the prominence of large screen TVs and the tendency for kids (and adults!) to excessively multitask, but we seem as far away as ever from the flying cars, weather control, and hover boards.  Curiously enough, the movie made no mention of areas where we have made enormous progress, such as computers that can beat chess and jeopardy champions, and consumer devices like iPhones and iPads.  Did Heinlein consult on this movie?

On the other hand, we've definitely arrived at the future in some aspects.  I realized we had crossed a line into the future when I saw my first web URL on a movie poster back in 1995 (it was for  Mortal Kombat).  If you grew up with the Internet, you have no idea how startling it was when this obscure bit of networking technology finally broke into the mainstream as people had been speculating it might for years.

Once it did, all bets were off.  All of us back then who were on the cutting edge (using advanced services like Compuserve and Prodigy) could have predicted email and Wikipedia.  Nobody could have predicted twitter as a force that could organize revolutions, blogging that would take on mainstream media, or youtube that would turn funny cat videos into global sensations.

So for purposes of figuring out when the future has arrived, I'm going to arbitrarily divide it into three basic stages:
1.    Computers and networks go mainstream
2.    Household robots take over household chores
3.    Flying cars

As noted, we're already well ensconced in Phase 1, and have a generation of kids and young adults who can't imagine the world any other way.  (And I must confess I still scratch my head when I try to remember how people used to find things out before Google, or even its predecessor Alta Vista.)

But if you think Phase 1 was a game changer, just wait until Phase 2.  This is going to revolutionize the very concept of what it means to be human, and will rock our society and economy to its core.  After all, if we have robots that can make the bed, do the dishes and take out the cat,
Do we still need housekeepers?
Do we still need cashiers?
Do we still need auto mechanics?
Do we still need airline pilots?

You might get a bit queasy thinking about that last one.  After all, do you really want to trust your life to a machine that might suffer a blue screen of death at any moment?  And you'd be justified in your concerns, as long as you ignore Bruce's Law of Technology.  Because shortly after we graduate from today's autopilot to something that can approximate a takeoff and landing in good conditions, the technology is going to  improve so quickly that insurance premiums for using human pilots instead of automated ones will go through the roof.  Expect automated pilots to catch on as fast as ebooks.

How far are we away from this transition to a robotic world?  It could be a long way off.  Decades.  In fact, from where we stand, it looks as far off as ebooks looked in 1986.  And 1996.  And 2006.

At Google's 2011 I/O Conference, Google announced an initiative to develop an open source operating system for robots (ROS).  As you might expect from a Google initiative, this OS will be easy to connect to the Cloud.  This means that robots can leverage Google's massive server farms for complex tasks such as object recognition, and not have to lug around the huge volume of CPUs that would be required to do this in real time, nor their associate batteries.

It also potentially means that researchers can share ideas, technology and databases much easier.  Why reinvent the complex algorithms to recognize a household object, when somebody else has already invested hundreds or thousands of hours on it?  Leverage their code, and spend your time thinking about what cool things your robot can do with that object, once it has recognized it.

Does this mean the robot revolution is just around the corner?  That's the problem with predictions - we just can't tell.  Google's latest initiative certainly looks promising.  But I could find other initiatives back in the nineties and eighties that looked just as promising.  The technology needed wasn't ready back then.  Maybe its not now, either.  Or maybe it is.  When it finally is, expect robots to turn our entire world upside down.  It will revolutionize life, work, employment, and leisure, although for good or ill is impossible to say.

All I can say is that humans better still be allowed to drive by the time I can finally buy my first flying car.  I'm sticking firm on that one.

No comments:

Post a Comment