Tuesday, August 12, 2014

Will computers replace writers? Nah, that's crazy.

IBM has created a new chip called the syNAPSE, a chip with more than 5 billion transistors, that models the human brain. For number crunching, this computer is slow. But for data pattern matching, like vision, sound and movement, this chip is built for the task.

Over the years, I've seen articles about attempts to replace writers with computers. This is, of course, a natural concern for anyone who wants to be a writer. So many industries are facing a competitive threat from computers. Even as I write this, a startup is preparing a machine to replace the fast food cook.

I saw this threat first hand in the sheet metal industry as I helped to inaugurate one of the first computer control plasma cutters in the HVAC industry. We had the first machine on the west coast and it was a wonder to behold as it cut metal with carbon dioxide, nitrogen gas and electricity. My supervisor became a river of unhappiness about the appearance of this machine and how many people it would put out of work. But I don't think that ever happened. When workers are displaced by automation, they find something else to do, often better than what they were doing before. Besides, someone has to tend to these machines until they can take care of themselves.

The Atlantic has an interesting story about a program that can take a set of data and turn it into a story - a readable, somewhat enjoyable story - if you're a sports fan. This reminds me of the Turing Test, a test suggested by the great Alan Turing. The Turing test says that at some point in time, we will create machines that can have a conversation with us and fool us into thinking we're talking to a human. You know, like the HAL9000 computer in 2001: A Space Odyssey.

Before I started writing this article, I spent some time over the last few days considering what a thinking, living, breathing human being puts into an article that a machine will never have to offer. The first thing to come to mind is instinct. Humans have a drive to live, they are self-aware and no matter what they do, self-preservation comes first. That is one very intense drive considering all the threats present on the planet.

There is something else that humans have that computers don't have: sensory perception. A computer will never see a sunrise in quite the same way that a human can. Computers are not born. They don't have any opportunity to gaze into Mom's eyes and identify with someone else. They can't have relationships with people because, no matter what anyone says, humans and computers will never be peers. There is nothing in our brains that will allow a computer to be a peer.

The last thing I want to bring up, and I think this is important, is this: quantum mechanics. Quantum mechanics (QM) is a set of rules we use to predict the behavior of the very small. It is also the best set of rules we have for predicting the way electrons behave in a circuit. QM is so accurate, we can use it to create chips that run our computers, crunch Pi to the billionth decimal place and beyond, and help us find our way in Google Maps.

Over the years, I've seen articles that document how life uses QM to live. The process of photosynthesis is well documented and shows how plants use QM to power life. All forms of metabolism use the rules of QM to work. I say this because it's easy to forget that what we see in a lab under controlled circumstances is something that is happening everywhere, all the time. The lab allows us to create conditions that make it possible to see this stuff happening. 

You might have heard of the famous double slit experiment. This experiment allows us to see light as a wave and particle. This experiment was created in a lab to show how light behaves in very controlled circumstances. Light acts as a wave and a particle all the time, but our senses are not built to notice that, at least not on a conscious level. The particle/wave duality of light is simply not germane to survival.

I think there is very good reason to believe that all life uses QM to live. I also think that QM is what makes our brains work and work well. Most of the time. We can still make mistakes and if we live past the mistake, we can learn from it. Maybe.

Until we can replicate what biology has done to make the brain, we will not be able to make a computer do what the brain does. The brains that we have now are the result of a billion years of trial and error, a direct result of the will to live. It's going to be a long, long time before we figure that part out. We might get close in the near term, but, at least in our lifetime, I don't think we'll ever know for sure how to make robots or computers self-aware. I'm not so sure we want to, either.

So for now, I think the writing profession is relatively safe from automation.

No comments: