By Michael Karasick
When Thomas J. Watson Sr. joined IBM in 1914 as its president, the firm didn’t have a single engineer on its payroll, so he quickly hired engineers and set up a product development group in a brownstone near New York’s Penn Station. He created a patent development department in 1932 and, in 1945, he established the first corporate scientific research laboratory. Today, IBM Research has grown to become the largest corporate research organization in the world, with 3000 professionals at 12 labs in 10 countries.
The point is that the nature of innovation keeps evolving and organizations have to change with it.
That’s why IBM is adopting a new approach to innovation for our newly formed IBM Watson Group, which will be headquartered in New York’s Silicon Alley. In the group, we are melding research, product development, experience design and collaboration with business partners and clients—all with the goal of accelerating the development of cognitive computing solutions for many of the world’s most vexing problems. This new era of computing requires a new approach to innovation.
Our Watson initiative builds on top of IBM’s long tradition of innovation, which placed IBM as the No. 1 recipient of US patents in 2013 for the 21st year in a row. We received 6,809 patents, easily outdistancing Samsung, the No. 2 finisher, with 4,676. The next US company on the top 10 list, Microsoft, ranked No. 5.
Among our 2013 patents, we racked up 250 within the cutting-edge cognitive computing category. That brings to about 1,400 the number of active cognitive patents we own. The new patents ranged from machine learning and natural language processing to neuromorphic computing and computer vision. Here are a couple of them that stand out as being fundamental building blocks for the new era:
Confidence assessor: This one came out of our work to boost the abilities of the core Watson question-answering technology—which won on the TV quiz show Jeopardy! It’s a method for automatically estimating the confidence the system has in answers it has come up with to questions posed by humans. The invention will help experts make better decisions. (Patent No. US 8,510,296 B2)
Synaptic computing: Modeled on the workings of the mammalian brain, this invention lays the foundation of a digital approach to spiking neural circuits through a combination of hardware and software design. The invention is being used in a new kind of microchip architecture designed to give computers sensory capabilities—including vision and hearing. You can anticipate its use in sensor networks and robots. (Patent No. US 8,515,885 B2)
I expect the IBM Watson Group to produce its share of patents, but its approach to invention will be different. In forming the group, we’re taking a team of scientists from IBM Research and combining them with software developers. My hope is that the experimental culture of the researchers will rub off on the developers, and that the get-stuff-done mindset of the developers will be adopted by the researchers. Also, integrating the two groups will reduce the organizational friction that sometimes impedes the progress of technology into the marketplace.
Meanwhile, IBM Research will keep pushing ahead with more fundamental inquiries into a wide range of cognitive technologies and science. As they produce breakthroughs, we’ll shift them into IBM Watson Group for commercialization.
But there’s more to the mix in the Watson group than combining research with development. We believe that an essential element of the era of cognitive computing will be producing machines that interact with humans in ways that are more natural to us. To achieve that goal, we need to add experience design and to experiment with real-world problems and data. First-class design will produce first-class user experiences.
So in the IBM Watson Group, we bring together people with expertise in user interaction—including experience designers, Web site designers, app designers and marketers focused on the millennial demographic. They collaborate to create new uses for Watson-style technologies and new ways for computing systems to interact with people. Independent application developers and our clients will also participate in these skunk works projects. I believe that if you mash together people with a wide variety of skills and knowledge, and you give them freedom to think as they will, interesting things will happen.
I’ve been at IBM for 25 years, and I have toggled from research to development twice before. This is the first time I’ve combined the two in one organization. So this is a new world for me—as it is for many of my colleagues. Our challenge is clear: We have to invent a new way of innovating so we can transform business and society in the new era of computing.
- Building on IBM’s Patent Leadership: Innovation in Watson-Style Computing
- Meet Another Person for a Smarter Planet: Master Inventor, Ed Suffern
- Why Patent Leadership Matters to Innovation & Economic Growth
- Linkin Park Harnesses Cloud Computing to RECHARGE the Planet
To learn more about the new era of computing, read Smart Machines: IBM’s Watson and the Era of Cognitive Systems.