Saturday, January 15, 2011

2011 - I gotta catchup

Hey there mathbaby,

You are asleep in the next room. It is 11:17 pm, and I don't get to sleep until 6am. It has been crazy/busy end of year (often is) and I have not posted. Now I'm going to working nights. Hope to pay off some debts, be able to have more time with you, be able to get some mileage in grad school. Will see how it goes. I miss you and your math-mommy already.

I read "Outliers" by Gladwell and they say look for big things. A big thing just happened. You need to know it, mathbaby. This defines the world you get to live in. If you understand it and make the most of it then you can have the opportunities of Bill Gates, or Steve Jobs.

This (link, link) shows the first computer that is better at human-only cognitive tasks than a human.

About 150 years ago we had a revolution called "the industrial revolution". It started small but the development of mechanical engineering made trains, and a very large variety of processing equipment - radically facilitating transport and production of goods, but also removing the employment of humans who had place in the former economy but not its descendant ones. Consequences of that revolution were machines that were stronger than (us) humans empowered us to go to the moon, to the bottom of the ocean, to dig deep in ...mountains, to make "fire" hotter than the sun. It transformed human ability and effort - and initially it was done wrong to high human cost. We nearly destroyed the world (think WWIII (yes that is 3), the cold war, and "75 overkill" - the multiple of number of times that current nuclear inventory could destroy the world of men), but gladly did not.

We are right now going over the edge of having computers (which have been faster than humans) able to consistently outperform humans and human-only intellectual tasks. It is likely that this will be the Gutenberg moment that I have been expecting since computing was born. This is a moment that redefines the way humans handle, not information, but understanding. Before Gutenberg people had to memorize books, afterward they could read them and it radically extended the productivity of all areas of human thought. Watson is an intelligence analyst right out the gate.

In 20 years, with the right kinds of economics and opportunities, it is likely that a pocket calculator will have the compute power of this "Watson" computer. Robots will have this level of performance - at least. Automated business trackers will have this level of performance. Automated accountants will have this level of performance. Automated lawyers will have (at least) this level of technical performance.
Automated snipers will have this level of performance.

This means that the only area that humans are best now only at being creative. We built robots to do dirty, dull, or dangerous - but it is likely that we will have, within appropriate contexts, relegated the majority of humanity there.

Someone is going to make an automated lawyer - it is going to start as a tool to help lawyers process information, then move up like mammogram analysis to something that counts as a peer to a human. After some time it is going to start replacing the humans for the low-risk and high volume work, divorces and wills, - because after all a real lawyer is expensive, and the computer is a beneficiary of Moore's law. It will ramp up to as high as it can consistently perform - and right now it is performing at better than Ken Jennings, the best human player of the game in the history of the game.

The problem with all of this is human history. Humans always ALWAYS try to destroy each other. Look at the nations in the cradles of civilization: the nile, the tigris/euphrates, the Indus, and the Yellow river. They show the best of what can be expected of 5000 to 10000 years of human history. Someone is going to make an automated accountant and fire a lot of accountants. Someone is going to make an automated lawyer. Someone is going to make an automated sniper robot.

The places with slower adoption are going to be places of culpability - you don't want to hold a computer culpable for loss of life, or property - but a human being can be liable. Robotic nurses will come much more quickly than robotic doctors. Culpability isn't the same as innovation. Innovation has already been moved in part to automated mechanisms that perform analytical processes at a masters-student level.

The easiest to imagine scary part (thanks Hollywood) is nation versus nation. If some nation makes the right sort of robot army then that army could seriously damage the ability of humans to not be extinct. The more likely and immediate is the days that created the unions - where the rich use robots as slaves and let the poor starve.

You, oh daughter of mine, get to live in the world halfway between this one and whatever they create whether it is utopia, or armageddon (or something in between).

If you live the MTV (slacker) model then you do nothing and get a future that someone else makes for you. Other people will often choose better for themselves and worse for you - don't let them do it. They will vote you off the island. I have seen too many bright (brilliant) young people transform their golden potential like a reverse Midas into a pile of manure and flush their lives irrevocably down the toilet. They never come back. Those opportunities never come twice.

If you live a hard-work and excellence model (I call it the Colin Powell model because it is a phrase from his autobiography) then you will be the one making the future. You can make it a win-win, balancing the madness of the century to come with a heart driven to bless as many people as possible with as much as you are capable of blessing. We are around for only a moment in the grand scheme of things .. our lives are a drop of water in the ocean of time. We have only this brief moment to do such great things, and we can - if we will but stretch forth our hand and try. Please be someone who tries.

How do I encourage you to be someone who will survive, thrive, excel? I wish I knew. Going to think about, and pray about it.

No comments: