Phil Neches was founder, chief scientist, and vice president of Teradata Corp., which applied parallel processing to relational databases. After Teradata was acquired by NCR, which was then acquired by AT&T, Neches served a number of roles for both companies. Joining George Gilder on the stage at WTF 2004, Neches spent some time comparing the current technology changes with those that occurred in the early 1900s. What follows is a partial transcript of his remarks.
The telcos are falling! The telcos are falling! Productivity is rising, wages are stagnating, jobs are going to foreigners, pricing power is eroding, mergers and consolidations abound, capital markets are confused, and the old ways don’t seem to be working. That sounds like the turn of the century. Not the turn of the 21st century, but the turn of the 20th century, the 1900s. Let’s look at that.
Let’s look at the transportation industry 100 years ago. The steel and railroad industries were in massive consolidation. The number of miles of railroad track in the U.S. was at its peak 100 years ago. It declined for the rest of the 20th century. Steel and rail started to lose employment. The steel industry mergers weren’t growth, they were consolidation. The automobile was an expensive toy. There were only 8,000 cars. There were only 144 miles of paved roads. And regulation? The maximum speed limit in most cities was 10 miles per hour.
Some fundamental things have changed. Steel is still with us. The interesting thing today is not steel per se but the things that we can make with it. Telecommunications existed 100 years ago — in about 8% of the homes. A three-minute call from Denver to New York cost $11, a rate some of our Bell heads would like to get back. And the idea that a regulated monopoly might be the best way to deliver this service was a new and controversial idea.
Let’s fast forward to today. Same principles. Life expectancy in 1900 was 47 years. 100 years later, we’ve increased that by more than half again. In the next 100 years, it’s not out of the question that we will do that again. 6% of all Americans had graduated from high school. These two industries, healthcare and education, are huge industries. Each is two to five times bigger than the telecommunications industry. They’re ripe for transformation, and thats going to take telecommunications technology. Lastly, crime? We used to think the Wild West was pretty wild, but compared to the average U.S. city, it was pretty mild.
In 1904, the average wage was about 22 cents an hour. Sugar cost 4 cents a pound. Eggs cost 15 cents a pound. But don’t forget the coal strikes. Talk about an energy crisis! It may be that the ultimate solution to the energy problem is controlling nuclear fusion through highly computationally intense control systems. IT could be the salvation of the energy problem.
Also, 100 years ago, marijuana, heroin, and morphine were not only legal, but you could buy them at the corner drug store. 14% of homes had a bathtub, and crossword puzzles and canned beer hadn’t even been created yet. If you spent all your time focusing on the problems of the time, you could’ve missed everything we consider part of the modern world.
I think we’re at the end of the third information era. The first era was mechanical processing and electric transmission. The second was electromechanical, the first half of the 20th century. The way we know now is electronic processing and photonic transmission. I think we’re on the eve of a new era, which is electrophotonic. But that’s not the end of the story. There will be a new new thing.
Some of the companies that were formed to bring early cash registers, calculators, and typewriters to market still exist. They’re still with us. The first telecommunication service was actually the fax machine. It didn’t succeed in changing everyday life in the way that later inventions like the telephone did because the mechanical technology of the time didn’t do a very good job until about the 1950s.
The second era was brought about by the fellow who figured out how to put an electric motor on a cash register. The other thing was the vacuum tube with gain. Having mechanical and electrical gain was all important because it made multi-stage processing possible. The equivalent of Moore’s Law then was at least a 35% rate.
Now today we have photonic gain. One of the things that’s made the electronic era a huge change that’s understandable was Moore’s Law. Electronics brought a 35% advancement rate, displays brought a 7% rate, and batteries brought maybe a 3% rate. The most expensive component in this laptop is the display, not the processor.
George Gilder: Displays are moving onto the Moore’s Law path, though.
Neches: Yes and no. Processors can get smaller and smaller and smaller. But displays are limited by human’s visual capacity. The need of the amount of glass in a display only increases as our eyes age.
Finally, there’s good old photonics, which is on a much faster learning curve. That’s setting us up for the next era. Soon we’ll be able to put photonic systems next to electrical systems on the same chip. And there are a number of other inventions such as carbon nanotubes, quantum computing, photonic switching, fluidics, and then some.
Look to the premier research universities in these fields. The possibilities they open up are just astounding. Nanotechnology is teaching us how to develop sensors and activators that can interact with our information systems’ senses. What sense do they have? A sense of touch, maybe through the keyboard. We’ve been trying to give them a sense of vision. But the real thing is that living systems evolved with chemical-based systems way before vision became important. One of the things thats going to happen is that the other senses will become important in information systems.
That makes for a fundamental paradigm change. It is cheaper to store and process a bit than it is to move them. In the electrophotonic era, that may not be true any more. It may become cheaper to move a bit than to process it. Everything we’ve learned about system design could be wrong. Right now we can only see “horseless carriage” applications. Everyone sees new technologies based on how it relates to or replaces existing technologies. New technology always enters the market at the margins. We still measure engine performance in horsepower. The big payoff of this technology is going to be something else.
We’re all feeling this set of discontinuities, this century change. But it’s happened before. And it’s happened before repeatedly. The thing that’s always carried us through is betting on humans’ ability to imagine. Imagination is the only recourse. And that gives me a sense of optimism, but that doesn’t mean that I’m going to hold onto my telco stock.