Costas J. Spanos

Costas J. Spanos is Andrew S. Grove Distinguished Professor, Electrical Engineering and Computer Sciences, University of California, Berkeley

Washington, DC – On May 10, 2018, the White House hosted the Artificial Intelligence for American Industry summit, to discuss the promise of AI and the policies we will need to realize that promise for the American people and maintain US leadership in the age of artificial intelligence.

The summit brought together over 100 senior government officials, technical experts from top academic institutions, heads of industrial research labs, and American business leaders who are adopting AI technologies to benefit their customers, workers, and shareholders.

To get an in-depth understanding of the Summit and its implications for the stakeholders, India America Today reached out to Costas J. Spanos, (More on him at the end of the interview below) Andrew S. Grove Distinguished Professor, Electrical Engineering and Computer Sciences, University of California, Berkeley. Spanos, one of the top experts among the attendees answered IAT questions, clarifying points for our readers.

After going through a summary of the Summit, provided by the White House, the following questions arise:

What is the present state of funding for AI research and what was achieved at the AI Summit?

The state of funding of academic AI research is relatively strong, and a very large component (at least at UC Berkeley) comes from corporations. To be clear, there contributors are interested in pre-competitve research, and they do not demand or expect that our findings would be shared exclusively with them, but rather published and disseminated freely.

Also, while this is happening in academia, we see very strong signals that the AI discovery work in industrial labs is accelerating, and a side effect of that is that a large number of academic faculty are being recruited away from academic into industrial AI labs.

The goal of the AI summit, as I understood it, was not focused solely on funding, even though participants gave a very strong signal that all parts of the discovery pipeline should be well resourced. This was especially true for the academic training and discovery, which needs to be well resourced in order to keep producing AI talent, which is currently rather scarce.

The initiatives under President Trump to liberalize AI innovation in transport, medicine, STEM, defense and other sectors have been announced. How far these will go to build AI in the US?

In my view this is a move in the right direction, but consumer protection (and building consumer trust) must still be a priority. We discussed at the summit the need to rationalize the distribution of the value driven from data, and that certainly means that the consumers who own their individual data must be protected form malicious use, and must gain a fair share of the benefits derived.

In the summing up the details, the White House noted it would be using “automation software to improve efficiency of government services and maximizing Federal data sharing with the American public, which will support non-Federal AI research applications.” How is that possible and how long will it take to develop such initiatives?

This is actually quite important. AI advance, discovery and applications demand access to large amounts of data, and, practically, this means that a few corporations today have a de-facto monopoly. De-identifying, curating and effectively sharing Federal Data will go a long way towards breaking that de-facto monopoly and democratize AI discoveries. This, however, despite the good intentions, will not be an oversight process, and will require some organization on the part of the recipients. For example, Research Institutes like CITRIS (Center for Information Technology Research in the Interest of Society – the one that I lead) will have to play a strong role distributing large datasets to individual researchers and to small research teams within academia. This is hard work and may take a few years to fully streamline it.

At the Summit, the White House noted, “To align interagency R&D priorities, we formed the Select Committee on Artificial Intelligence.” What do you see the role of this committee in taking AI initiatives forward?

As the name implies this is meant to coordinate the actions of the various agencies. This is crucial, since many agencies have grown a bit organically, so at present there is discontinuity and redundancy. This is definite a step in the right direction, and I would be curious to see how it will operate.

Any other points that you think are important for our readers to understand?

Yes. A key point is that there is a difference between how AI researchers understand this technology and how the public perceives it. Beyond the hype, researchers see huge possibilities and challenges, and one of the most critical challenges is that AI technologies to-date are not adequately “explainably” and thus not “trustworthy”. I believe that a key objective should be the development of a technical/legal/policy framework to enhance and track the “trust” that we should have on AI technologies in critical applications (Health, Defense, Transportation, etc.)

Thank you for your time and effort in helping our readers understand the Artificial Intelligence Summit.

More About Professor Costas J. Spanos:

Professor Spanos received the EE Diploma from the National Technical University of Athens, Greece, and the M.S. and Ph.D. degrees in ECE from Carnegie Mellon University. In 1988 he joined the department of Electrical Engineering and Computer Sciences (EECS) at the University of California, Berkeley, where he is now the Andrew S. Grove Distinguished Professor and the Director of the Center for Information Technology Research in the Interest of Society (CITRIS), a Research Institute across UC Berkeley, UC Davis, UC Merced and UC Santa Cruz.

Professor Spanos is also the Founding Director and CEO of the Berkeley Education Alliance for Research in Singapore (BEARS), and the Lead Investigator of a large research program on smart buildings based in California and Singapore. Prior to that, Professor Spanos has been the Chair of EECS at UC Berkeley, the Associate Dean for Research in the College of Engineering at UC Berkeley, and the Director of the UC Berkeley Microfabrication Laboratory.

Professor Spanos’ research focuses on Sensing, Data Analytics, Modeling and Machine Learning, with broad applications in semiconductor technologies, and cyber-physical systems. He has authored or co-authored more than 300 papers, 15 patents, and a textbook. He has supervised more than 40 Ph.D. recipients who now hold key positions in industry and academia. In 2000 he was elected Fellow of the Institute of Electrical and Electronic Engineers for contributions and leadership in semiconductor manufacturing.


Please enter your comment!
Please enter your name here