Judy Estrin is the most influential Silicon Valley woman you’ve probably never heard of. A Los Angeles native and the daughter of two UCLA computer scientists, Estrin, a UCLA graduate herself, later studied at Stanford under the legendary Vint Cerf, developer of the Internet’s popular protocol, TCP/IP. As a scientist and entrepreneur in Silicon Valley, Estrin co-founded three companies that had successful exits: two IPOs (one later acquired by 3Com), the other an acquisition by Cisco. Estrin later served as Cisco’s chief technical officer and also as a board member at FedEx and Disney. In 2000 she wrote a paper that helped popularize today’s ubiquitous term “cloud computing.”
Q: What did you have in mind when you wrote your famous “Clouds vs. Strings” paper in 2001?
Estrin: We were advocating for the new converged voice/data infrastructure to look more like the Internet than traditional telephony. Today the cloud has an expanded definition. It means massive computational power available “in the cloud,” i.e. over the Internet.
Q: How did the cloud’s rapid evolution occur--from voice over IP, which you wrote about, to the ubiquitous powerful computing it provides today?
Estrin: It goes back to IP becoming the choice for data, voice and video, thus contributing to more and more enterprises being comfortable with the Internet as the fundamental means of connectivity. It enabled cloud services to begin to be offered to customers that could pay for it.
Q: Your hiking pal Diane Greene, who cofounded VMware and now runs Google Cloud, said something astonishing at our Forbes CIO conference in April. She believes the underlying pace of evolution in the cloud is two to three times faster than Moore’s Law.
Estrin: I don’t have specific numbers on it, but there’s no question that things are changing. The accelerating pace of technological change is a phenomenon we all need to be very, very aware of. I think there are a lot of factors that play into this. As we’ve moved more computing to the cloud, it takes less time to adopt technological changes because the cloud vendors do a lot of the integration and each one of those clouds are serving multiple customers. It used to be if you had a new application, a new technology or a new way of doing something you sold it to each enterprise. Well, if enterprises and small businesses are essentially using the same cloud by Amazon, Microsoft and Google, then change in either the server technology or the networking bandwidth, or in the applications that run on top of them can move more quickly.
Q: Over the last two years there’s been an explosion in the use of such terms as “artificial intelligence,” “machine learning” and “robotics.” Is that because the rapidly accelerating cloud is the engine that drives these?
Estrin: When you look at machine learning and how it’s being applied in cloud environments, there’s no question that the cloud is one factor in the acceleration. But it’s really a combination of technologies--the processing in our devices, whether it’s phones, computers or sensors, combined with the level of consolidated computational power in the cloud-based ecosystems--that allows this steeper curve of more processing power at cheaper costs.
Q: For all the excitement about the cloud, when I interviewed Michael Dell earlier this year, he pointed out that the cloud is still only somewhere between a 15% and 20% share of enterprise computing and storage today. Does the cloud eat everything, or will there still be a market for on-premises computing and storage?
Estrin: That seems low to me. When you look at the impact of cloud computing, I think you need to look at it not just in terms of the enterprises directly served, but that there are so many companies providing either consumer or enterprise applications, whose applications are hosted on the cloud. So the cloud is actually serving a broader set of consumers. You know, often we’ll use a service and we don’t even know what cloud it’s on. It’s not just enterprises that are impacted by this, so I think we shouldn’t underestimate the impact.
Q: Let me put it another way. Are we approaching a time when everything goes to the cloud?
Estrin: I think we’ll see more and more of that, though there will always be some applications and enterprises that choose for a variety of reasons not to be cloud-based. It may be control, it may be management, or it may be a perception of more security. And those larger enterprise organizations that use private cloud services may continue to have some systems that they don’t want moved into the public cloud.
But I’ll also throw out the notion that we’ve seen this pendulum swing a number of times between highly centralized and fully distributed systems. Every time we’re in the middle of a swing we assume we’ll never go back in the other direction. I’m not saying we will. I think today’s cloud architecture is being supported by cost-efficiency, among other benefits. But we don’t know if the pendulum will go back to the middle or swing further in 15 to 20 years. We can’t underestimate the fact that things get disrupted, even when we think we’re going to have a system for life.
Q: How should CEOs think about cloud technology?
Estrin: CEOs must think about the cloud. One benefit is that it accelerates the adoption of new technology, which means it also accelerates business evolution. Your competitors are thinking about it; you can’t just delegate to your tech people.
Q: Yet even today most corporate boards ghettoize information technology to the audit committee. Is that a mistake?
Estrin: Yes. Boards need to consider how technology drives strategic advantage. That’s how Fred Smith and FedEx saw it when I served on FedEx’s board and chaired its IT Oversight Committee. Fred was ahead of his time.
Q: Speaking of FedEx, you were its first woman director.
Estrin: I was the only woman and considerably younger than the rest of the board. I was 36 when Fred asked me to join and had built my career in Silicon Valley in a very different way, doing startups. It was hard at first. I did a lot of listening and found ways in which I could contribute. I built relationships.
Q: You served on FedEx board’s for 21 years--and on Disney’s for 15 years.
Estrin: Disney was a very different company when I joined. Michael Eisner was the CEO. He and Disney had just gotten written up for having the worst board in America because it was too insular. Disney was just figuring out the Internet, an area in which I had experience.
Q: When you were sitting on the FedEx and Disney boards, how much time did you allocate to that work?
Estrin: It varied. I probably spent more hours on Disney than on FedEx, partly because there was more going on at Disney--the Comcast bid and the Eisner-to-Iger succession. Board seats aren’t full-time jobs, but you always have to be prepared and ready to jump in with both feet at any time.
Q: What advice do you have for people who want to join boards?
Estrin: All companies are being impacted by the accelerating change of technology. Boardrooms are crying out for directors who not only understand technology but also understand how to navigate through change.
Q: Paradoxically, often the best thing a board can do, as you’ve said, is to slow things down.
Estrin: Asking the right questions of management is key, especially in a very reactive world. For an athlete to be agile, he or she needs to have a strong core. Knowing what not to react to is also important. You can’t run after every change.
Q: Every board says it wants more women. How should they go about recruiting more women?
Estrin: The same way they look for any member--seeking excellence and someone who is compatible with and complements the group. They’ll end up with better board members if their approach is, “Yes, we’re interested in diversity. But we’re also interested in these types of capacities or domain expertise, and in this type of leader.”
Q: Silicon Valley, your home turf, is still pretty hostile to women on boards. What’s going on?
Estrin: Board members too often come from venture firms or are former CEOs. The resulting problem isn’t just the sexual harassment you’ve been reading about, it’s an insularity that hurts agility and dynamism. We need to broaden the field we’re drawing from.
