Interview with Dario Gil, worldwide director of IBM Research Labs

17/01/2023 1

Darío Gil is one of those Spanish scientists who emigrated to the other side of the Atlantic in search of success. In this case, professional success. He started, the first six years, working in the semiconductor world, especially in lithography and chip density. He was the first to realize the technology of immersion lithography, where water was put between the lens and the silicon wafer. Years later, he became, as he puts it, “an R&D manager working in various areas such as physical sciences, and data analytics for smart grids”, … which led him to lead the entire Science and Technology area of the company. He was also responsible for Artificial Intelligence, and it was then that he started with the whole issue of Quantum Computing, where he progressed until three years ago he was appointed worldwide Director of IBM Research. And now he is also Senior Vice President and Chief Technology Officer at IBM, and head of all intellectual property globally. Translated with www.DeepL.com/Translator (free version)

“There doesn’t seem to be the same concern in Europe and the United States for safeguarding data privacy.”

He has also been involved in the world of Science and Technology Governance in the United States. He was a member of the White House Science and Technology Committee, and now he is also a member of the National Science Board, which is the Commission responsible (24 people) for the National Science Society, which supports basic science in the United States with an annual budget of some 10 billion dollars. “I am also very involved in science policy and its areas of investment,” he says before we begin to ask him.

We talked with him about the Governance of Science and Technology in Spain, about quantum computing today and tomorrow, with forecasts that they will become increasingly powerful to solve industrial, business and social problems with systems that next year will have more than 400 qbits, or in 2023 more than 1000.

And Dario Gil also told ConfIA by Sherpa.ai how quantum computing will improve the development of Artificial Intelligence. “One of the great directions of artificial intelligence is to create very complex neural networks that learn from data without having to formalize symbolically, ex ante, what it is that we are trying to learn. A new field is emerging, which is quantum machine learning, and Artificial Intelligence techniques can also be used to improve quantum systems themselves. For example, in assessing the fidelity or measurement of the quantum processor itself. Using AI to classify and emulate what is happening when building a system.

One of the great directions of artificial intelligence is to create very complex neural networks that learn from data without having to formalize symbolically, ex ante, what it is we are trying to learn.

Very interesting were his reflections when we asked him about Data Privacy, a field in which Sherpa.ai is working on its new data privacy platform based on Artificial Intelligence. Although he agrees with the strict term of privacy, and the non-sharing of data, understanding as very positive the use of platforms such as those developed by Sherpa.ai, explains Dario Gil. And when we talk about data privacy, he points out that “Europe has a very sophisticated and advanced vision in the regulatory field of advanced technologies. And the United States has a more practical orientation. If Europe wants to be a leading technological power in Artificial Intelligence it cannot make its regulation the only prism. It should also have the spirit of being avant-garde, pioneering, … and not see everything with the lens that scientists want to take advantage of Society.”

There are many elements where there are constitutional protections, “and beyond the more protected rights there are laws and regulations that we all have to comply with. There is already a whole regulatory context that influences how we have to use technology. What happens is that when new challenges or advances are generated we have to reinterpret existing laws to cover new scenarios of use”, concludes Gil, leaving the door open to new developments such as those developed by Sherpa.ai, and always respecting privacy and the non-sharing of data.