The Singularity: Apocalypse or Nerd Rapture?
In the Terminator series, John Connor and his mother fought to prevent Judgement Day, the moment when the computer system Skynet becomes self aware, decides humans are a threat and moves to wipe us out.
In the real world, we have our own version of Judgement Day called the Singularity and many can’t wait for it to come. Like in the Terminator it is the moment when computers take over, but rather than killing us off the machines enhance us.
Some have called the Singularity a nerd rapture. Its prophet is Ray Kurzweil, one of the world’s foremost experts on artificial intelligence and the new Director of Engineering at Google, where he will direct a good portion of the company’s $6.7 billion R&D budget to making his vision come true. Like it or not, science fiction is becoming science fact.
The Rise of the Robots
In February 2011, the game show Jeopardy! held a special competition. As per usual, there were three contestants. The difference was that one of them was not a human, but IBM’s Watson, a supercomputer.
Despite the format of the game, in which intuition is as important as intelligence, Watson won handily. Ken Jennings, one of the contestants and the show’s all-time money winner, wrote as his final answer, “I, for one, welcome our new computer overlords.” Pretty soon, all of us will feel that same sense of excitement mixed with foreboding.
Maniacal robots aren’t chasing us through city streets yet, but they’ve already started taking our jobs. We’re beginning to experience a new productivity paradox. Whereas before economists worried that investments in information technology weren’t paying off, now the fear is that they are becoming so efficient that there will be no role for humans.
The notion may seem fantastic, but the transformation is already underway. While robots have been a fixture in factories for decades, they have now started taking over highly skilled human jobs, such as legal research and creative jobs in arts and culture once thought soley in the human domain. IBM is even sending Watson to medical school.
When asked what was the most powerful force in the universe, Albert Einstein is reported to have said, “compound interest.” While the quote may be apocryphal, it is a similar force at work that produces accelerating returns.
By now, most of us have become aware of the digital laws that double the power of our technology roughly every 18 months. In ten years, our devices will be 100 times more powerful, 10,000 times more powerful in 20 years and a million times more powerful in 30 years.
Perhaps not surprisingly, we don’t use computers to do the same things a million times faster than we did a generation ago, but to do completely new things. The effect can be summarized in this cartoon that Kurzweil included in his 2005 book, The Singularity Is Near.
We’ve come to expect the unimaginable to become commonplace. Some of the tasks still on the wall, such as driving cars, reviewing movies and translating speech, have already been taken off in the short time since the book was published.
Yet what’s really driving our race to toward the singularity is the fact that the accelerating returns phenomenon is now permeating areas which we have not ordinarily associated with computers, such as genomics, energy and manufacturing. Kurzweil predicts that in the future, “all technologies will essentially become information technologies.”
An Emerging Form of Life
In truth, nobody has been able to come up with an acceptable definition for life. Webster’s dictionary says it’s “an organismic state characterized by capacity for metabolism, growth, reaction to stimuli, and reproduction,” while the biologist Richard Dawkins regards living beings as vehicles for code to replicate itself.
The scary thing is that whatever standard you use, it’s getting harder and harder to argue that the machines we are creating don’t meet it. In fact, the technology we are building closely mimics our own functions:
Decision Makers: Computers have long been designed to make basic decisions, which at a fundamental level, is the function of an algorithm. However, newer techniques, such as Bayesian nets, can learn from bad ones and get smarter over time just as a child does.
Pattern Recognizers: Kurzweil was a pioneer in creating hierarchical hidden Markov models (a technique for deep learning), which form the basis for how computers can understand our handwriting and speech. As he points out in his new book, How to Create a Mind, scientists are discovering that our brains work much the same way.
Evolvers: Darwinian evolution is, in fact, a fairly simple algorithm. Different entities compete under certain conditions and the best ones survive into the next round, where new variations are introduced and the process starts all over again. Similar genetic algorithms are widely used in logistics systems and other areas in which a particular solution needs to be optimized.
The mathematician Samuel Arbesman predicts that as our computers continue to get smarter and are applied to ever increasing data sets, they’ll begin to make discoveries that we can’t even understand, much less uncover ourselves.
Self-Replicators: While our machines still need us to create them, it’s already becoming clear that there will come a day when they won’t. Eric Drexler has been theorizing about self-replicating nanobots since the ‘80’s. We’re not there yet, but the technology is advancing at a rapid clip.
There is an old notion, called the infinite monkey theorem, which states that if you had enough monkeys tapping mindlessly at keyboards, producing works of genius such as the novels of Tolstoy or Shakespeare would be a function of curation rather than creation.
As the power of our technology continues to advance at an exponential pace and processing becomes, for all intents and purposes, limitless and free, we’re starting to experience a real-life version of the infinite monkey theorem.
Computers are beginning to outperform us not because they are always right, but because failure costs them virtually nothing. They are able to simulate millions of variations and their outcomes before taking action.
The Rise of The Cyborgs
The Terminator was, of course, a work of fiction, but the idea of cyborgs and killer machines is not all that farfetched. In fact, the danger is becoming so real that the Obama Administration found the need to spell out explicit rules that specify under what circumstances machines are allowed to kill humans.
The line between human and machine capabilities and actions is getting increasingly blurry. Many of us, me included, find it hard to function without our smartphones, which link us to all the world’s information from basic facts to where to find a decent burger. Google’s Glass project can now put the same functionality in front of our eyes.
It’s not hard to see how we all might be rushing out to buy technological implants in a few years and that’s the Singularity’s ultimate joke. There is indeed good reason to fear the rise of the cyborgs. After all, they’re us.
Who said the universe didn’t have a sense of humor?