I LEAD a team of scientists that is designing a cognitive computing chip and software ecosystem inspired by the human brain. It will help computers sense, perceive, act and reason more like humans. In the face of much skepticism and tremendous technical unknowns, our goal is finally within reach.
In the not-too-distant future, I envision chips based on this design powering everything from robots that rescue people after natural disasters, to talking eyeglasses that help blind people navigate through tricky situations, to cars that are more aware of the driver and the road.
This was no easy journey. It involved coordinating the activities of a range of scientists from my company and other organizations. We had to work in virtually uncharted territory under the real-world constraints of a relatively short deadline and limited funds.
The project, called Synapse, is financed by the Defense Advanced Research Projects Agency, or Darpa. In various project phases, the team has included world-class, multidisciplinary scientists from major universities, government agencies and I.B.M. labs.
From the start, the team was guided by the notion that good ideas can come from anywhere or anyone. We also adopted a handful of core principles to guide our progress. Early on, we decided to pursue multiple technical paths, with a predetermined timeline and metrics. As a group, and in concert with Darpa, we evaluated our progress and decided which paths to follow.
The project faced an existential challenge when some of our initial basic assumptions proved unworkable. Certain options would have taken too long and cost too much to meet our deadlines and metrics.
At one point, it looked as if the project would fail. But we worked through the obstacles and came up with a new solution. The key insight was to shift focus from inventing new technologies approximating the brain's neurons and synapses to inventing a new computer architecture with existing chip technology, which can then serve as a beacon for future technology development. This decision required letting go of long-cherished ideals while embracing emerging possibilities.
At a decisive moment, I called a meeting of all the participants. About 25 people crowded into a room, and others were on the phone as part of a brainstorming exercise called "six thinking hats." This technique, pioneered by the author and inventor Edward de Bono, requires a debate's participants to identify, by using colors, the essential nature of the points made -- white for facts, black for discernment, red for emotions, green for creativity, yellow for optimism and blue for coordination.
During our encounter, everyone was heard and various team members took turns jotting down the "colors" of the statements that they and other people were making. In this way, the participants could see more objectively whether their arguments were driven by facts and solid reasoning, and whether they were being open-minded.
The technique lightened the mood and let us resolve our differences without making the debate too emotional or personal. More important, it allowed us to tap into creative possibilities. Ultimately, we reached a consensus and set off on a new direction that led to a successful project. As more challenges unfolded, we repeatedly returned to the same blueprint.
Here are the most important takeaways from my experience:
• Identify problems and confront them proactively.
• Under deadline pressure, use a flat management structure, allowing people to contribute ideas rapidly.
• To sustain morale and keep team members engaged and productive, share credit for successes and learn from failures.
• Remember that the creative potential of committed individuals, if harnessed properly, has great potential to bring constructive change.
This experience has taught me valuable management lessons. In the coming years, it seems likely that organizations of all kinds, from the private and public sectors, will need to collaborate with one another in order to succeed. Research and other aspects of business will globalize even more, and new paths to innovation will require people from different disciplines to work together to combat uncertainty while harnessing potential. The key is to channel a group of superstars so that all are pulling in the same direction.
Dharmendra S. Modha is a senior manager of cognitive computing for I.B.M. Research.employment
This article originally appeared in The New York Times. First Published October 12, 2013 2:01 PM