Leading the Future of Technology
By 2030, scientists and engineers will have the ability to create an entirely new class of supercomputer, predicts IBM Senior Vice President and Director of Research Darío Gil ’98, one that combines the best of quantum computing, AI and classical computing. The greatest opportunity with this new supercomputer: the acceleration of scientific discovery, Gil says.
“Why does that matter? I don’t think that we have run out of problems to solve in the world,” he says.
This supercomputer could greatly compress the time it takes to discover solutions to profound challenges, Gil says, from building better batteries to power the world, to creating a better type of fertilizer to feed it.
Gil is leading the effort to create this supercomputer —and whatever’s next in computing — as he oversees the global research strategy for the corporation’s more than 3,000 researchers.
Gil, who is based out of IBM’s research headquarters at the Thomas J. Watson Research Center in Yorktown Heights, New York, leads research strategy for AI, hybrid cloud, semiconductors, quantum computing and exploratory science. He co-chairs the MIT/Watson AI Lab, which is advancing fundamental AI research. Gil is an internationally recognized technology expert, chairing the National Science Board, which oversees the National Science Foundation. Last year, he appeared on “60 Minutes,” discussing the tremendous capabilities of quantum computing and offering a sneak peek of what was then the soon-to-be launched Quantum System Two.
This past summer, Gil sat inside IBM’s Think Lab, surrounded by Quantum System Two – IBM’s latest quantum computer, operating at temperatures colder than deep space. Nearby was the Quantum System One, the company’s earlier supercomputer.
Quantum’s great strength lies in its potential to do certain calculations much faster and much more accurately than classical computers, he says. IBM was the first company to build a programmable quantum computer and make it available through the cloud. Since then, it has built more than 80 quantum computers, for clients including the Cleveland Clinic.
Also displayed in the Think Lab is the Artificial Intelligence Unit (AIU) cluster, a glass-enclosed chamber that powers IBM’s watsonx AI system for corporate clients and organizations.
Applying AI to computer programming is going to be transformative, Gil says, with the biggest impact on scientific discovery. Having AI that can read research papers, tables and charts — infinitely more material and more quickly than any human could — and generating hypotheses from these will be game-changing, he predicts.
But the most profound development in the future of computing, Gil says, is creating a whole new class of supercomputers that seamlessly integrate the power of quantum, AI and traditional computers. At IBM, Quantum System Two is the platform for this future technology.
“Our ultimate goal is to create a quantum-centric supercomputer that is going to bring the best of classical computers, the best of AI and the best of high-precision computation, with the best of quantum, all working together,” he says. “And that’s what we want to deliver by the end of the decade.”
Gil speaks passionately of the great potential of quantum and AI. He acknowledges being deeply frustrated by the state of the debate on AI starting in 2023, when hyperbole about its potential dangers hit a boiling point, he says. A group of scientists and researchers signed letters pretty much predicting that AI could lead to the end of the world. He calls it ridiculous.
“It created a level of alarm in society that was an appropriate response, given what was being said,” he says. “It created a challenging environment.” More recently, he sees things cooling down a bit.
“I think there’s so much positive that AI can deliver in terms of productivity, economic growth and problem solving, but we have to manage it, we have to govern it,” he says. “And that means that we have to take responsibility for the people who create the tech, and responsibility for what use cases are appropriate and which are not.”
Dissatisfied with the level of debate, he created the AI Alliance, which has brought together 110 institutions — companies, universities, science agencies, non-profits — that work together to shape the future of AI. One shared belief: AI should be open and transparent about how data is used and how the technology is trained.
Gil was an international student from Madrid, Spain, who transferred to Stevens as a junior, drawn to the university by its proximity to New York City. Stevens gave him his first research experience — in tribology, the study of friction at the nanoscale level — with Professor Traugott Fischer. Gil loved the small scale of the campus that allowed him to get to know his professors, who guided him on his path to Massachusetts Institute of Technology, where he earned a Ph.D. in electrical engineering and computer science. He graduated as the Stevens Class of ’98 valedictorian.
“I thought it was wonderful, and it made a big difference,” he says of his Stevens experience.
Decades as a leader and researcher — he has spent his career at IBM, joining as a researcher in 2003 — have led to insights he shares with those launching their careers. Collaboration and clarity of communication with colleagues is vitally important, he says.
“It opens lots of doors,” he says. “Lots of other people can help you, and you can help them, to achieve things you couldn’t do by yourself.”
In research, trends come and go, but there’s no substitute for depth of knowledge in an area, Gil says. Knowing something deeply allows you to recognize nuance and the limitations of your own understanding.
“You know what you don’t know, and you ask better questions. You can have more comfort in being humble on things that you don’t know,” he says.
When Gil thinks of those who have influenced him, he thinks of his family, many of them creative and artistic, with an independent streak, who never sought the validation of others. He thinks of mentors at Stevens, MIT and IBM, who taught him to not only be an engineer and researcher but also to pay attention to detail and “the rigor of things.”
“I have had the great privilege of learning over time, and people teaching me the complexity of things and the nuance of things. That has greatly influenced me,” he says.
“There is a great joy in practicing your field, year after year, and decade after decade. It never gets old,” he later adds.
As someone who is constantly thinking about technology and its future, Gil insists it’s all about people.
“I am not a technological determinist,” he says. “Technology is a very powerful tool to help us solve problems faster and better than we could do by ourselves. We need to set the bounds on what we want and what we don’t want from technology.
“It’s all about people in the end.”
– Beth Kissinger
Darío Gil was the featured speaker at Stevens’ President’s Distinguished Lecture Series in October.