Ray Kurzweil

Ray Kurzweil was the opening act at ASTD TechKnowledge this morning. His message is awesome. Once you grasp his Law of Acelerating Returns, you can’t help but change your view of the world. In a nutshell, not only is progress advancing exponentially, the rate of change itself is exploding exponentially, too.

Ray plots the time between paradigm shifts from the beginning of life to the present day. Plotted on a logarithmic scale, it’s a straight line. We’ve been evolving exponentially faster since the beginning of time. We hadn’t noticed before because we were in the flat part of the curve; now we’re at the knee. Soon we’ll be on a rocket ride, where everything becomes a blur.

I’m going to post a longer entry than usual below, because I think Ray is a pragmatic visionary who has it right. There’s no time to lose.

Before he spoke, I told Ray I loved his new book, The Singularity is Near, but that its intimidating size (it’s a couple of inches thick) is bound to hurt sales. He told me he felt he had to show all the evidence he had collected to support his argument.

“I was at Bill Gates’ house not long ago talking with Bill and some senior guys about eLearning.” The technology is here; it works; it is available. eLearning will be a major focus of the Gates Foundation. Thousands of schools are offering the MIT Open Courseware; 60% of MIT’s course lectures are now online.

Information technology grows exponentially. We are learning a lot about how the brain works (it is another information processor.) Moore’s Law applies to all aspects of life. Size is shrinking. Decentralized info technology is toppling C&C. “I was talking with Mikael Gorbachev at lunch…” and Gorby confirmed that the dissemination of information is what toppled the USSR.

Exponential trends sneak up on you. We are now at the knee of the curve, where we can actually see explosive growth. The rate of technological progress is exponential. Most of today’s jobs did not exist a century ago.

We are now doubling the rate of paradigm shift every year. The 21st century will contain about 20,000 years as we used to know them. Phone: took half a century to reach half the population. Radio, Television, PC, mobile phone, the Web are coming on faster and faster.

The Law of Accelerating Returns is actually a form of evolution. Ray puts up a logarithmic graph that plots time intervals between paradigm shifts, the pace of evolution. The graph is a straight line, from the dawn of life through multicellular organisms, the Cambrian Explosion, reptiles, mammals, primates, spoken language, Homo sapiens, art, early cities, agriculture, writing, the wheel, city-states, printing, the experimental method, industrial revolution, phone/electricity/computer, and to today’s PC. Ray also ran the numbers of the evolutionary steps as described by Carl Sagan, Britannica, and a dozen others; once again, a straight line signifying exponential growth.

The throughput of information technologies of all kinds doubles every year. Computers have doubled in power 24 times since the 1940s. Won’t technology hit a wall (like rabbits in Australia)? Moore’s Law is projected to tank in 2020. No, there is always a new paradigm, from cards to relays to tubes to transistors to chips. When chips hit the wall, three-dimensional computing will replace them.

The price of a transistor has been cut in half every 1.5 years since the late sixties. From $1 each to millions of transistors for a dollar. The line is nearly perfectly straight.

Each of us runs on 23,000 little software programs called genes. Many are running obsolete code. It is already possible to turn off genes that lead to obesity, atherosclerosis, etc. Biology has become an information technology. Instead of “discovering” drugs, we’ll engineer them for specific purposes.

We think we live in a linear world. The web feels like it came out of nowhere, but actually it was in the flat part of the exponential curve.

Brain holds 30 to 100 million bytes, less than the code of Microsoft Word. Human intelligence is based on pattern recognition: the quintessential example of self-organization. (I missed something here because Ray chops the brain down to size the way a gif file reduces the size of a bit-mapped image: by representing a thousand like elements not with a thousand bytes but with 1000x one of them.)

2010: computers disappear. 2029: $1000 will buy a computer with 1000x the power of the human brain.

Nanobots. With full VR, you could swap bodies. Or live forever. The life expectancy of a Cro Magnon was 18. At the dawn of the Industrial Age, people lived to the age of 37. Now we’re up to 70. Computer intelligence will surpass human intelligence within the lifetimes of people alive today. If we can upload our minds, we could lead disembodied lives.

At some point, Ray’s predictions begin to sound like science fiction. He didn’t say it, but he intends to be immortal, uploading his wetware when his body gives out.

Ray’s forecasts ring true for me. They change my perception of time. As life goes into fast-forward mode, rigid structures break down. Books become obsolete before they hit the shelves. College credentials lose all meaning. Jobs can no longer be described. Knowledge becomes fluid. Life requires perpetual adaptation. Unless we speed up the clock-time of our brains by manipulating our DNA, the only path to adaptation is learning. The faster the changes, the more learning will be required.

There won’t be time for learning as we have known it, for the curriculum would be changing day to day. (And before long, minute to minute.)

Northern Californians know what happens to brick buildings in earthquakes. Shake them up and cracks bring them down. By contrast, wooden buildings flex with the strain and ride things out. San Francisco’s redwood Victorian houses survived the quake of 1906 because they were able to go with the flow.

Whenever I come upon a rigid learning structure, for example a lock-step technical training program or the graduation requirements for engineers, I want to shore it up by inserting flexibility. In the software world, this is called “loose coupling.”

The internet is nearly indestructible because it is loosely coupled. It is made up of “small pieces, loosely joined.” It routes around failure. When threatened by seismic change, the net flexs. This is a good model for learning to adopt. It doesn’t ask for permission to adapt to new conditions. There’s no boss there. It trusts in the power of self-organization. For managers, the lesson is to take control by giving control.

That’s one of the major factors that has converted me into such a fan of informal learning.