NEW YORK – The Hellenic-American Cultural Foundation (HACF) presented Professor Constantinos Daskalakis: ‘When will machines be truly intelligent’ on October 12 at Merkin Concert Hall at the Kaufman Music Center in Manhattan.

Daskalakis, an award-winning MIT professor and computer scientist, discussed the state of Artificial Intelligence (Al). While AI has made great advances in changing many facets of our lives, it is also becoming increasingly clear that AI poses significant threats.

It is also known that AI often fails to work well in novel environments and may amplify stereotypes and biases that exist in the data used to train it. Prof. Daskalakis took a critical look at the developments in AI from a combined scientific and applied perspective, revealing intimate connections to mathematics and social sciences as potential avenues to overcome these challenges.

He also discussed the new Archimedes Center for Research in Artificial Intelligence in Greece and how it plans to advance AI research.

Daskalakis began his lecture by offering some background on the history of computing, noting that there were machines in the past that performed specific computations, like the astrolabe, but later the tremendous contributions of mathematician Alan Turing, who is perhaps best known as the character portrayed by actor Benedict Cumberbatch in the film The Imitation Game, paved the way for the programmable computer. Daskalakis also mentioned Ada Lovelace, one of the first programmers, the daughter of Lord Byron, who abandoned her in England while he went off to help Greece during the War of Independence.

He pointed out that “general purpose computing allows us to do remarkable things,” such as the moon landing, but also to simulate physics that allows for the design of atomic bombs and other types of bombs, so computing like every tool can be used in a good way or in a bad way. “The same is true for AI,” Daskalakis said.

He then explained what AI is, citing three definitions including Astro Teller’s “AI is the science of how to get machines to do the things they do in movies,” the Wikipedia definition – “an intelligent agent is a system that perceives its environment and takes actions which maximize its chances of success,” and his colleague Patrick Winston’s “Artificial Intelligence is the study of the computations that make it possible to perceive, reason, and act.”

The professor then noted the approaches used for the development of algorithms for image and speech recognition, highlighting the successes and the limitations. For example, there is an algorithm that can ‘paint’ photos in the style of a famous artist’s painting such as Van Gogh’s Starry Night. A byproduct of predicting the next word in a sequence for another algorithm is the ability to correct English.

Daskalakis noted, however, that “AI systems are unreliable,” they can be manipulated and are biased. He shared amusing examples of completed sentences gone awry but also mentioned more serious ones, such as how AI is used in the criminal justice system to help determine who is released from jail pre-trial. Daskalakis also shared a list of what is important for developing reliable AI including sample size and the correct use of statistics.

Daskalakis then spoke about the Archimedes Center for Research in Artificial Intelligence, Data Science and Algorithms, founded in January 2022 with the support of the Committee Greece 2021, funded for its first four years by the EU Recovery and Resilience Facility (RRF), and operating as a Research Unit of the Athena Research Center. The Archimedes team includes Daskalakis and Christos H. Papadimitriou as Principal Scientists and Timos Sellis as Scientific Director. The vision for the center is to bring Greece to the forefront of research in AI and also build bridges with diaspora scientists.

The lecture was followed by a Q&A session and a reception.

Christina Hioureas, member of the HACF Board of Directors, gave the welcoming remarks at the event and introduced Daskalakis, a long-time friend of hers from their days at the University of California, Berekley, though they were in different departments. Hioureas noted that Daskalakis’ parents were both born in Crete, but he was born and raised in Athens where he attended the Polytechnic, earning his degree in Electrical and Computer Engineering. Daskalakis went on to study at the University of California, Berkeley, earning his PhD in Electrical Engineering and Computer Science in 2008.

Between 2008 and 2009, he was a postdoctoral researcher at Microsoft Research, New England, and has been at MIT since 2009. He became a tenured Professor in 2015 and full Professor in 2018. Daskalakis works on computation theory and its interface with game theory, economics, probability theory, statistics and machine learning. His work on the complexity of the Nash equilibrium, with Paul Goldberg and Christos Papadimitriou, received the Kalai Prize from the Game Theory Society and the Outstanding Paper Prize from the Society for Industrial and Applied Mathematics (SIAM). He has also received the 2008 Doctoral Dissertation Award from the Association for Computing Machinery (ACM), the 2010 Sloan Foundation fellowship, the 2011 Spira Award for Distinguished Teaching, the 2012 Microsoft Research Faculty Fellowship, and the 2017 Google Faculty Research Award. In 2018, Daskalakis was awarded the prestigious Nevanlinna Prize by the International Mathematical Union for “transforming our understanding of the computational complexity of fundamental problems in markets, auctions, equilibria, and other economic structures,” and the prestigious Simons Foundation Investigator award in Theoretical Computer Science, an award designed for “outstanding scientists in their most productive years,” who are “providing leadership to the field.” In 2019, he was awarded the ACM Grace Murray Hopper Award and the Bodossaki Foundation Distinguished Young Scientists Award.

More information about the Archimedes Center is available online: https://www.athenarc.gr/en/archimedes.