I recently spoke with Silicon Valley veteran Bill Coleman about Silicon Valley and its future. Mr Coleman worked at VisiCorp on the first spreadsheet, he headed development of Sun's Solaris, he co-founded BEA Systems which sold to Oracle for $3.5bn, and he now runs Cassatt, a cloud computing tech startup.
I'm sitting with Bill Coleman, and one of his colleagues, in a small conference room in downtown SF. We're talking and also trying to eat our sandwiches.
He is telling a story about being at JFK airport late last year, and getting a call from a reporter at the New York Times, asking if the recession would be better or worse for Silicon Valley compared with 2001.
"I said it will be worse than the previous downturn. The last time about 800 companies went out of business that had no business model. Secondly, the large companies had sold too much capacity and it took a year or two for things to catch up. This time the recession is not about Silicon Valley it is about something else."
He said it would take a long time for capital spending to come back and that this would hurt Silicon Valley.
"I said you'll see accelerating layoffs in the first quarter and that's exactly what we've seen."
What will this mean for Silicon Valley innovation?
Mr Coleman says he's an optimist but he does have some concerns.
"I have a basic theory that Silicon Valley reinvents itself by inventing a new platform layer every 10 years." He says that Silicon Valley was lucky to develop Information Technology (IT), a technology that is becoming cloud computing, a new platform. Information Technology is also vital in driving the development of two additional disruptive technologies: nanotech and biotech. And fortunately, Silicon Valley leads in all three industries.
Silicon Valley also leads in green technology, a large and growing market. But green technology is different -- it isn't a disruptive technology. He says that a disruptive technology has to have a characteristic of the Peter Drucker rule in that it provides ten times the value of what it's displacing.
"In the green market, what we're displacing is cheaper per unit than what is displacing it. It won't be driven by a tsunami of adoption." Mr Coleman says the government should come up with incentives for companies to adopt green technologies otherwise progress will be slow.
"People already have the capacity for doing what they do today. So this means that they can put off using green technologies for a long time. There are lots of benefits for humanity, but the economics of the green market won't drive a rapid adoption unless there are incentives."
He points out that a government program focused on incentives to adopt green technologies would provide a more effective stimulus than the current stimulus package, with its focus on building physical infrastructure. Building a bridge, or a road, is a one-time event, and it won't provide long term stimulus. "It slows down capital formation," he says, potentially slowing an economic recovery. If the government helped to expand green technology it would create higher quality jobs and provide other long term economic benefits.
Cloud computing doesn't need government incentives because it is a disruptive technology, says Mr Coleman, especially the next stage, beyond what he terms "cloud 1.0." As the cloud computing platform becomes more sophisticated, he predicts that there will be an acceleration in the use of the cloud driven by a "quadruple conversion." Video, audio, and IT data all become IP based, and productivity applications become integrated with social networks.
"As we move forward from Web 2.0 to Web 3.0, all your productivity tools become integrated with your social networking, which becomes your business networking. Your mobile life and your online life will become the same. So now the client moves into the cloud and that's when we'll see a dramatic change in the cost structure of computing and of the capabilities you can have."
He says that a platform ill be successful if it has three characteristics. It has to be able to commoditize a market. Secondly, it has to obey the 10x better/cheaper rule.
"When I was at VisiCorp, heading software development, each of our developers had a PC. This was faster and dramatically cheaper than using DEC VAXs [minicomputers]. Thirdly, a platform must allow you to add value with custom additions. The reason Netscape wasn't a platform was that no one could program to it, nobody could add value. (By the way, that's also true for virtualization...) Unless you have all three characteristics, you won't have a disruptive chain that can accelerate a startup from zero to sixty, and turn it into a major player."
I mentioned that a characteristic of a disruptive platform, is that it disrupts. If you are in the path of a disruptive technology you can often see what's up ahead, you can see the train wreck, but you can't get off the rails in time, you can't downsize or restructure in time, you hit the train wreck.
"Yes. I call it the DEC spiral. DEC tried to deny the importance of the PC and then when they realized what was happening they couldn't layoff people fast enough to match falling prices. There was little left and they had to sell [in 1998 to Compaq, the top PC company at the time.]" said Mr Coleman.
(Please come back for the second part of this interview...)