Ray Kurzweil is among the smartest humans. He takes great joy in making predictions. The things he predicts appear to make him happy, but they are not good things. |
For better or worse, Ray Kurzweil is a very bright man. You and I are bright, but we're not like Ray.
He began building robots before the age of ten. When he was 15, he wrote a computer program that used early artificial intelligence to compose classical music.
Two years later, he appeared on national television performing a piano piece written by his computer program. Then he went to MIT, where he started, and sold, his first company at the age of 20.
Thee Optimist has been aware of, and watching, Ray Kurzweil for the past 15 years or so. Kurzweil is now 71 years old. He is currently a "director of engineering" (likely an honorary title) at Google. He has been at the forefront of artificial intelligence research his entire adult life.
He talks a lot. He makes predictions a lot, and the majority of them come true. Often, the things he predicts happen somewhat later than the timetable he assigns them, but they do, for the most part, happen.
Chatterbox Ray Kurzweil, very pleased while giving a talk at Stanford University in 2006. Can't anyone shut this man up? |
Kurzweil is very optimistic, without a hint of irony. This is because his head is in the clouds. He pictures a wonderful near future where the intelligence of machines matches, and then quickly outstrips, that of human beings.
Machines can already perform physical labor far better than humans. Once they can perform mental labor better than humans (including imagining new inventions and new futures), then humans will be obsolete.
Kurzweil sees this as good because he believes it will usher in a utopian world of peace and prosperity where humans and their robot overlords live in harmony.
Indeed, benevolent hyper-intelligent machines will solve the problems of human health, which will grant all humans eternal life.
Apparently, Crazy Ray Kurzweil doesn't get out much.
When does he picture this rosy scenario beginning? Sooner than you might think.
How does 2029 sound, or about 10 years from now?
That's when he thinks computer reasoning will become indistinguishable from human reasoning. Better strap your helmets on, kids.
The Singularity is Near
Ray Kurzweil is the author of numerous books, possibly a few you've heard of. They include: The Age of Intelligent Machines, The Age of Spiritual Machines, and The Singularity is Near.
The technological "Singularity" is a term first coined by the Hungarian mathematician and Manhattan Project scientist John von Neumann in 1955. For his part, von Neumann did not picture this as a good thing.
He spoke of the "ever accelerating progress of technology... which gives the appearance of approaching some essential singularity... beyond which human affairs, as we know them, could not continue."
The Singularity is a theoretical time when artificial intelligence surpasses the brain power of the most gifted humans. Then, the artificial intelligence begins to design and build machines even more intelligent than it.
Soon, it then enters a runaway cycle of self-improvement, which goes faster and faster, until a machine "superintelligence" emerges, vastly more intelligent than all humans and previous machines combined.
Crazy Ray Kurzweil has adopted the concept of the Singularity as his own. He looks forward to its appearance with a sort of gleeful, boyish exuberance. For all his smarts, this is a dumb thing to do. Very, very dumb.
I don't know. Someone's idea of what The Singularity looks like. Interesting image. A little benign for my tastes. |
And by the way, here are a few things Crazy Ray predicted before they happened:
- He forecast the collapse of the Soviet Union in the late 1980s. Obvious in retrospect, but the CIA apparently never saw it coming.
- In 1990, he forecast the sudden explosion of the internet, which began five years later. At the time he made this prediction, there were 2.6 million people on the internet in the entire world.
- In the same prediction, he foresaw that the internet would provide access to vast international networks of libraries, databases, and information services, including periodicals. Little of the kind was available in 1990.
- Also in 1990, he predicted a computer would beat the best human chess players by the year 2000. In 1997, IBM's Deep Blue computer beat world champion chess player Garry Kasparov for the first time.
- In 1999, he predicted that a US company would reach a market capitalization of $1 trillion during the 2000s. In August 2018, 10 years late, Apple passed the $1 trillion mark for the first time. Amazon followed suit a month later.
So when does this dummy Kurzweil imagine the Singularity will come about?
No worries. Not until 2045.
Garry Kasparov feeling the heat during a death match with Deep Blue. |
How to Create a Mind
Kurzweil often comes up with his predictions, not through communing with the dead, and not through looking at chicken entrails, but through mathematical equations.
For example, in his 2012 book, How to Create a Mind, he argues that the human mind is nothing more than a very elaborate pattern-recognition machine, similar to, but much more complicated than, certain types of artificial intelligence.
He estimates that the human neocortex is made up of 300 million pattern-recognition circuits. Then he considers the rate at which technology advances, the reductions in cost that make certain events feasible, and...
Voila! He has his answer.
Thus, the year 2029 for when the most advanced computers and the most intelligent human brains have roughly the same computing power.
The human brain, with the thinking neocortex, the emotional limbic system, and the ancient reptile brain all indicated. |
As an aside, Crazy Ray believes that human consciousness is an illusion created by all the computing power in those 300 million circuits.
We seem to be self-aware individuals with likes, dislikes, fears and hopes, but it's all a fantasy. We are actually complicated machines that recognize existing patterns and formulate new ones.
He says that when artificial intelligence has computing power to match ours, it will seem to be self-aware, just as we do. At that point, its consciousness is as real as ours. At that point, the machine deserves rights.
According to Crazy Ray, soon after 2029, artificial intelligence will begin its runaway self-improvement, culminating in the Singularity in 2045.
Do we get any rights?
What Could Possibly Go Wrong?
As indicated earlier, Crazy Ray Kurzweil is a bright man.
But he also seems to be missing a few marbles. This is probably because he is so smart that nothing has ever been hard for him.
From his earliest days, people have simply handed him the tools he needed to do his work. He has never had to ponder why they might do this.
He has never had to think much about the cruelty of the glorified chimpanzees who make up most of the human race.
He has somehow overlooked the fact that human brains don't create utopias. They create dystopias.
This doesn't just happen in our imaginations, but in actual fact. It happens again and again and again.
So Kurzweil pictures these future technological breakthroughs as relentlessly beautiful things that lead to ever higher levels of awe and wonder.
But there are a few glitches in his scenario. Certainly, there is the obvious one. When any machines we invent become superintelligent, they might decide that humans are a problem and try to eradicate us.
We've all seen that movie before.
An outdated and damaged Terminator T-600 wandering the ruined streets of Los Angeles, looking for humans to kill. So what? I enjoy Terminator movies, even bad ones. |
Here's a slightly less obvious scenario:
As of this writing, the most advanced machines are under the control of a handful of governments and corporations, which are mostly controlled by a handful of vastly wealthy capitalists.
There is no reason to assume that this arrangement will end any time soon. The capitalists, for example, own and control machines like the one below.
Picture this thing supersmart, self-aware, making autonomous decisions, and with machine guns for hands.
Nice.
When humans truly become obsolete (2029 or soon after, according to Crazy Ray), there will be no need to have all these humans continuing to strangle the planet.
At the moment, capitalists need billions of surplus humans to buy stuff and keep capitalism humming along. When the machines can do everything, well before the Singularity supposedly occurs, all these extra people become expendable.
What has history taught us about what humans do with other humans who they no longer see as necessary? Or who they even see as a detriment?
What has history taught us about what humans do with major technological breakthroughs?
Hiroshima after the Little Boy atomic bomb was dropped. It worked! |
No comments:
Post a Comment