The Computer Scientist who Challenged the Nash Equilibrium

In November 2014, I was privileged to meet Constantinos Daskalakis, a Greek Computer Scientist and Associate Professor at MIT’s Electrical Engineering and Computer Science department. Daskalakis is best known for his work in the field of Game Theory, and his research on the Nash Equilibrium has earned him prestigious awards and an international reputation.

2014-11-18 11.54.32

Constantinos Daskalakis and Mina Theofilatou, November 2014

I had contacted Constantinos by e-mail earlier in 2014 regarding the awareness campaign Greek CS Teachers were running against the insane decision of the former Education Ministry’s administration to eliminate Computer Science from the National University Entry curriculum (read more about the problem here and about how it was resolved here) and, as one would expect, he was very supportive of our cause. When I wrote to him after booking my air travel for the CSTA November Board meeting – I had a 12-hour layover in Boston on the return leg, perfect for a visit to MIT! – he was willing to make some time in his hectic schedule to meet with me.

The meeting took place in his office at CSAIL in the Ray and Maria Stata Center; I had never been to Boston or MIT before so I followed the instructions he gave me (“it’s a crazy Frank Gehry building, you can’t miss it!,” he wrote). Our conversation naturally started with the situation in Greece, and inevitably ended with a discussion on the future for young computer scientists. I asked what he would advise exceptional students wishing to further their education at establishments such as MIT. He mentioned that what he is mostly looking for in Ph.D. applications is evidence of potential for ground-breaking research; that the clearest form of such evidence is prior research engagement; and that recommendation letters are crucial in assessing the student’s research potential.

That’s sound advice for those pursuing post-graduate studies, perhaps even an academic career. But what about Computer Science in Primary and Secondary Education? I asked Constantinos if he could write a few words about the importance of students having early exposure to computational thinking and computer science principles. Here is what he had to say:

“‘Information’ and ‘computation’ are just such fundamental concepts that there is no doubt they should be an integral part of primary and secondary education. All that takes place around us (or inside us) can be viewed as computing on information. Sometimes we want to process information efficiently—think Google trying to rank webpages. Other times we want to mine interesting information from vast amounts of raw information—think trying to identify chunks of the genome implicated in some disease. Sometimes we want to hide information—think cryptography. Other times we may want to release information while respecting privacy of individuals—think releasing medical data. Some other times, we want to incentivize individuals who have information to reveal that information to us…. And it is not just us trying to gain or operate on information. Biological, physical, and social processes are fundamentally computational, operating on information in some application-specific medium. Computer science principles can change one’s perspective on life, science, and society. Our educational system should teach that way of thinking early on.” 

And a summary of how his work has impacted international knowledge in computer science?

“My research studies the foundations of Economics from a computational standpoint. The starting point is that, besides sometimes being irrational, humans are definitely computationally bounded – in some sense, irrationality is a form of computational boundedness. So, studying economic behavior needs to incorporate computational thinking into economic thought. My research focuses at this interaction between Economics and Computation. One of my most celebrated results was showing that the Nash equilibrium – the crown jewel of prediction tools in Game Theory, defined by John Nash, the mathematician portrayed in “Beautiful Mind” – is computationally intractable. This means that, in complex interactions, it may take centuries of time before humans (even if they have access to super-computers to help them with their computation) behave as Nash equilibrium predicts. This casts doubt on how accurate the prediction ability of Nash equilibrium is, calling for better, computationally aware tools for predicting behavior. Research at the interface of Computation and Economics hopes to shed light on such issues of fundamental importance for understanding human behavior.”

A big thank you to Constantinos for providing valuable insight into the importance of Computer Science at all levels of education.

Mina Theofilatou, CSTA International Representative, Kefalonia, Greece