Science fiction has been around a very long time, but it wasn’t always called that. The essential core of science fiction has always been three promises: space travel, intelligent alien beings and intelligent robots. We know as far back as the classical Greece, that there has been speculation about travel to other worlds and finding intelligent beings on them. The idea of building an artificial human is as old as memory too.
There’s always been a few outliers in society that think up far out ideas and a larger group of fans who favor them. Currently we call these two groups science fiction writers and science fiction fans. During the second half of the 20th century I believe certain science fiction ideas peaked in popularity, and that we’re now detecting a possible diminishment of their popularity.
I strongly felt the public turning against the major promises of science fiction when I read the new issue of The Atlantic, and the essay “The Man Who Would Teach Machines to Think” by James Somers that profiles Douglas Hofstadter, author of the 1980 Pulitzer Prize winning Gödel, Escher, Bach. Hofstadter hit a home run with his first book, but has been mostly missing in action all these years since, even though he continues to write brilliant books about artificial intelligence (AI). The trouble, as The Atlantic article points out, is Hofstadter’s idea of artificial intelligence is different from what the academic world has come to accept for the term. Douglas Hofstadter wants to teach machines to think, just like us, while the industry is happy enough to program computers to accomplish fantastic data processing feats that give the illusion of thinking.
If we want robots like C-3PO, then we need Douglas Hofstadter. If you’re happy with IBM’s Watson, then we don’t. And I’m worried that most people on Earth don’t have the sense of wonder that it takes to want a C-3PO. And that’s a fucking crying shame. I want Star Trek, but the public is only grudgingly willing to pay for NASA. I want humanity to become friends with all the aliens in the galaxy, but all the vast run of people on Earth want is to thrill to the xenophobia of alien invasion movies and to shoot ETs in video games.
I have pretty much given up on seeing the public support space travel, and figure our only hope of meeting aliens would be through SETI projects, but I figured we had a real chance of seeing intelligent machines in my lifetime. I might have to give up on that dream too.
For most people artificial intelligence is not an issue they will ever concern themselves with, but if you’re a philosopher, computer scientist, or science fiction fan, then it is. The crux of the matter is whether or not machines will ever be able to think like us. Here’s my logic. Humans are self-aware thinkers and we’re the accidental creation of evolution. If nature can randomly rub molecules together until it produces a self-aware biological being why shouldn’t we create thinking machines intentionally? Sure, it took 13.78 billion years for reality to create us, but that doesn’t mean it will take as long for us to engineer intelligent machines if we put our minds to the task – we have 13.73 billion years of experience to consciously study.
Up to now, we’ve mostly tried to program machines to do specific jobs, some of these jobs used to be tasks we thought required thinking, like playing chess, being a contestant on Jeopardy or translating foreign languages. We can program machines to do these tasks, but they don’t think, not in the way we think. That’s not a failure of AI, it’s a lesson in what makes us conscious beings.
To do what Hofstadter wants will require building machines that can learn and evolve. This is completely different from the direction that AI is taking now. We can’t program machines to be self-aware, but we should be able to program machines to learn and evolve, and eventually that will lead to self-aware AI.
Think about the evolution of life on Earth. It reflects the growth of simplicity into complexity. It shows how simple creatures learn to interact with its environment and evolve better senses. Over time those senses could interpret more and more complex patterns in the environment. Look around you. Everything you see is recognize as a distinct object. In a cluttered room you might be seeing hundreds of different things. Think how, and how long it took you to learn what all those things are. Computer scientists for the longest time have tried to just tell machines what to see. That won’t work for a thinking machine. Like a human child, a thinking machine will have to grow up and learn everything on its own.
It does no good to create code that tells a computer what a banana is. Can you remember learning what a banana was, and how to tell it from all the other kinds of fruits, or even distinguish it from vegetables? I bet you can remember learning what an iPhone is, and maybe you can even tell the difference between an iPhone 3S and a 5S. You’d think it would be easy to tell a computer to do the same thing, but it’s not. Modern AI can be programmed to spot an iPhone, but not out of context of knowing what everything else is around it. Not seeing and understanding the complete context of the visual field shows why the machine isn’t thinking. It’s how we learn about new things that’s thinking, not knowing what they are.
The same problem we face building thinking machines are the ones we face for creating true space travel and finding alien life forms in the galaxy. Most people just don’t see the point. They don’t want to waste the money. And they’re xenophobic. But what it comes down to is most people really don’t care. It’s not on their radar. Space travel, aliens and robots have no value to them at all. Zip. Nada. Nothing.
So why the immense popularity of science fiction at the movies, on television and in video games? Well, that’s another essay.
When I was a kid back in the 1950s and 1960s I embraced science fiction because I wanted to see space travel in my lifetime. I wanted first contact in my lifetime, even if it was just SETI contact. And I expected intelligent machines to be created in my lifetime. Hell, I thought all of these things would have happened by the beginning of the 21st century. Boy, was I wrong.
I thought we took a bad turn when the Apollo program was cancelled, but felt things were back on schedule with the emerging popularity of Star Wars and the return of Stat Trek. I felt millions and millions of Earthlings were embracing the three great promises of science fiction as science fiction at the movies became huge box office successes. But I was wrong. All those people weren’t dreaming the same dreams as I had. They are getting something else out of science fiction.
Orphans of the Sky is a powerful story by Robert A. Heinlein about a generation ship traveling the vast distances between the stars for so long that the passengers have forgotten that they live in a spaceship. Their self-contained world becomes their entire universe and they forget the larger universe exists at all. When I read Orphans of the Sky back in the 1960s I felt that humanity was waking up and realizing that they were living on spaceship Earth. Hell, again I was so goddamn wrong.
In most people’s minds, people on Earth live in a small place where God rules over them and cruelly holds out promise of everlasting life if they only confess belief. Earth is a ball of dust God created as a classroom for us to live on while we decide. Earth has no purpose other than a staging area for heaven and hell. All the rest of the vast universe is a big distracting illusion.
The future used to be a vision of mankind spreading out to the stars making endless discoveries, but now I have a different vision. Humans will continue to live on the Earth forgetting it’s a spaceship traveling through a vast universe, and the inhabitants will continue to follow their illusions century after century until they destroy their ship. I could be wrong – I’ve shown that to often be true. Let’s hope. Maybe The Enlightenment is just taking longer than we thought.
Now it might sound like I’m depressed over this reevaluation of my beliefs, but I’m not. I consider it far more healthy to be realistic than try to keep my own cherished fantasies. The truth is always discovering the true mission of the spaceship where you become self-aware. Which brings me back to robots. If we ever create truly self-aware robots, what will they make of old spaceship Earth?
JWH – 11/6/13
Auxiliary Memory 
“But what it comes down to is most people really don’t care. It’s not on their radar. Space travel, aliens and robots have no value to them at all. Zip. Nada. Nothing.”
That statement sums it up. Most people are more concerned about the Kardashians or who’s playing each other on Sunday than things that could truly better their world. We live in what I like to call “connected ignorance” in many ways today. That is, while we’re more connected than ever before, our collective intelligence is seemingly in decline. Critical thinking is sparse and the idea that humans can, and should, seek out advancement and truth is rejected by all but a few.
Your synopsis regarding the future of humanity is likely true but never give up your ideals because of the lack of conviction held by others. Being a realist is a good thing but do keep in mind that it’s often the dreamers that truly change the world for the better. Good post Jim.
Realism is always handy, so here’s my take:
1. Space travel has, and is, happening. Not to the extent I might have though reading SF in the 1960’s, but it’s does exist. What we lack are, first, much better propulsion systems to reduce travel time by orders of magnitude, and, second, a way to use space travel to replenish the resources used to make it happen. It doesn’t have to produce a profit for rich people in the old-fashioned capitalist sense, just deliver a payoff that’s accepted as a useful tradeoff.
2. The universe could be full of intelligent beings, but if the speed -of-light barrier imposed by Einstein turns out to be unbreakable, then the chances of our meeting any of them are tiny. So, sorry, Dr. Fermi, there is no paradox. We’re all just too far apart. (Unless, of course, our counterparts are living on planets very near Earth. That we haven’t been visited by them is probably down to the fact that they haven’t been visited by us. I.e., neither of us knows how to make that kind of trip.)
3. Machines get cleverer and more capable all the time and each advance pushes out the AI boundary a bit more. We just aren’t comfortable with putting the intelligence tag on a machine we built that operates in ways we predetermined. I.e., if humans wrote a robot’s source code, humans aren’t likely to call it intelligent. Very likely we won’t until a machine that looks like something from the cover of an Asimov novel moves in next door.
All that said, the contemporary long-running fixation on dystopias and failures in science fiction reflects a sad lowering of expectations and a cynical lowering of standards that is liable to be self-fulfilling.
I’m not sure its the fault of science fiction – sure there’s plenty of downer SF, but for the most part, SF is full of hope for the future. What disappoints me is most of our fellow humans just don’t care about the exciting futures that science fiction imagines.
I think you need to write a novel about the slow process of teaching computers to think. I think that’s one of the things science fiction can do in some circumstances: Make people more used to an idea.
That’s in fact, what I’m doing John.