CEO & CO-FOUNDER, NUMENTA
Donna Dubinsky is CEO, board chair, and co-founder of Numenta, Inc., a software company creating fundamental technology designed to lead the coming era of machine intelligence. Donna first partnered with Jeff Hawkins (co-founder of Numenta) at Palm, Inc. in 1992, where she served as president and CEO. She held this position throughout Palm’s acquisition by U.S Robotics and subsequently 3Com Corporation. In 1998, Donna and Jeff co-founded Handspring, creator of the category-defining Treo smartphone. Handspring merged with Palm in 2003, and Donna continued to serve on Palm’s board until 2009. Previously, Donna spent 10 years at Apple Inc. in a multitude of sales, sales support, and logistics functions—both at Apple and at Claris, an Apple software subsidiary. Donna earned a B.A. from Yale University, and an M.B.A. from Harvard Business School. She is currently serving on the board of Yale University as the Senior Trustee.
I entered the computer industry when it was very young by joining Apple Computer in 1981. I had been inspired by seeing a demonstration of an Apple II running VisiCalc, and I was convinced (rightly, as it turns out) that personal computers would have a huge future. Since then, I’ve had a front row seat at four major computing revolutions: personal computers, handheld computers, mobile computing and now, intelligent computing. At my current company, Numenta, we are studying how the brain works and how brain principles will be used in machine intelligence.
We are very excited about the potential for building intelligent machines modeled after the human brain. Although there is extensive work going on in artificial intelligence that is solving problems today, we believe that true breakthroughs will come from biologically inspired computing. Today’s AI operates on very different principles than does your brain, so there is much to be learned from human intelligence.
Our biggest challenge at Numenta is a scientific one: figuring out how the brain works. This involves in-depth neuroscience research and a focus on large-scale brain theory and simulation. We like to say that we are at the intersection of computer science and neuroscience. While there are many companies interested in building intelligent machines, we don’t know of any others that are taking this unique approach.
In terms of opportunities, I see several. As mentioned earlier, I’ve been on the pioneering side of four major computing revolutions. In each case, when we built the platform, we couldn’t imagine the type of applications that would result. For example, when we created the PalmPilot, we never imagined Instagram. When we created the Treo smartphone, the first commercially successful smartphone, we never imagined Uber. Today, working on machine intelligence, we also cannot imagine what the most exciting applications will turn out to be.
When we started Numenta in 2005, we had a clear mission: to lead the new era of machine intelligence by learning how the brain works. Over the years, we have gone through several changes – different business models, leadership, and applications. Yet while the details have changed, the mission has remained constant. Staying true to our mission helps guide our decisions about the business.
I fully expect that the work we are doing will be reflected in a variety of commercial products in the coming decades. We are in the business of licensing intellectual property and our hope is that this approach will enable a wide variety of solutions to be deployed.
We are focused today principally on advancing our research. We have several major papers under development that we think will inspire the community to think more creatively about how to enable a computer to solve problems more like a brain.
The most important attributes to be successful at business are flexibility and resilience. You almost never end up with the business turning out to be the way you imagined it. You have to be constantly questioning your assumptions and changing when appropriate.
Our major frustration at Numenta is that everything takes longer than we would like. We’ve made huge progress in our research, and more important work is underway, but we always wish it could be faster.
Because we are licensing intellectual property and not offering services or products to customers, we strive to be as open as possible with our work. Transparency and education are key. We openly publish our scientific findings, software, intellectual property, and business strategy. There are no hidden agendas. A few years ago, we even made the decision to open source our daily research code. While other companies might look at this decision as a risk, we believe the benefits of conducting our research in the open outweigh any drawbacks. We have an active open source community, and this gives them visibility to our priorities, helps them understand the work more deeply, and facilitates collaboration. It also enables people to develop their own applications in their specific domain.
We believe that the work we are doing is creating fundamental new ideas that can positively impact our society. While many companies are focused on near-term projects that are highly focused on solving a specific problem, we have a more general approach and are excited by providing an impact that could be widespread well into the future. Consequently, some people are very attracted to this type of long term, substantive mission.
I would encourage people starting out today who are interested in machine intelligence to read some of our work and experiment with some of our technology. I expect that the next generation of technologists will be able to apply this work in many interesting ways. Just as I was able to build a successful and impactful career by getting on board with personal computing in its infancy, young people today should consider getting on board with biologically inspired computing.