SEPTEMBER/OCTOBER 2006 (Vol. 10, No. 5) pp. 6-8
1089-7801/06/$31.00 © 2006 IEEE
Published by the IEEE Computer Society
Published by the IEEE Computer Society
Judea Pearl Interview: A Giant of Artificial Intelligence Takes on All-Too-Real Hatred
PDFs Require Adobe Acrobat
Judea Pearl has reaped many awards throughout his distinguished career. Pearl, 69, a professor of computer science at the University of California, Los Angeles (UCLA), is a fellow of both the IEEE and the American Association of Artificial Intelligence. In 2003, he received the Allen Newell Award from the ACM for his lifetime contributions to computer science, notably in the fields of Bayesian networks and artificial intelligence.
In awarding Pearl the Newell prize, the ACM said, "He forged links between computer science and statistics, developing models that are used to describe everything from the effects of diseases to the likely behavior of terrorists. Pearl's ideas have revolutionized the understanding of causality in statistics, psychology, medicine, and the social sciences. His book, Probabilistic Reasoning in Intelligent Systems [Morgan Kaufmann, 1988], is among the most influential works in shaping the theory and practice of knowledge-based systems."
In a profile in UCLA Engineer, one of Pearl's colleagues, professor Richard Korf, called him "one of the giants in the field of artificial intelligence," and also characterized Pearl as "very single-minded in focusing on his research — sort of the classic researcher at a university. All else was a distraction."
Pearl is currently nominated for yet another prestigious prize — the inaugural US$100,000 Purpose Prize, sponsored by the Atlantic Philanthropies and the John Templeton Foundation. The prizes will be awarded to five Americans over the age of 60 who are making extraordinary efforts in the area of social innovation. However, Pearl's nomination for the Purpose Prize isn't connected to his formidable contributions to computer science and networking; instead, the nomination is rooted in a personal tragedy, the 2002 kidnapping and murder of his son, journalist Daniel Pearl.
Shortly after Daniel's murder, the family founded the Daniel Pearl Foundation ( www.danielpearl.org), dedicated to creating better cross-cultural communication. One of the Foundation's projects, the Daniel Pearl Dialogue for Muslim-Jewish Understanding, is a publicly staged conversation confronting the necessity to address the grievances held by both Muslim and Jewish communities worldwide. Joining Pearl in the dialogue is Akbar Ahmed, professor of Islamic Studies at American University. The dialogues have been conducted in the US, Canada, and the UK.
The two are nominated for the Purpose Prize as a team because of their dialog's impact on its listeners. Pearl recently shared his views with IEEE Internet Computing on the Internet's worldwide utility, for both good and ill — one of the world's foremost theoretical researchers compelled by personal circumstance to ask network engineers and managers to think about how the Internet is used to spread messages of both hate and hope, and how to help address that conundrum.
IEEE Internet Computing: Prior to your son's death, you were best known as a mathematician and artificial intelligence expert. Since you founded the Foundation, have you done any more theoretical mathematics?
Judea Pearl: Sure, sure. I keep my office at UCLA. I am a "professor on recall," teaching one class a year, and I have a cadre of PhD students who have obtained some nice results lately, and that's the main thing.
IC: Do you consider any work you've done lately to be potentially new and groundbreaking?
Pearl: Yes. We have obtained a breakthrough recently with my student Ilya Shpitser, proving that the calculus of intervention that I introduced in 1995 is complete. This means that any valid conclusion can be inferred by applying the three rules of inference I posed.
IC: Do you know of any networking companies that are exploring Bayesian networks? Microsoft, for example, was doing some research in them.
Pearl: Microsoft has a whole team working on Bayesian networks. They started with diagnosis, helping users troubleshoot systems using the least amount of technical intervention. When Bill Gates came to UCLA to give a lecture, someone asked him what Microsoft was doing for the long-term future. He answered, "Ah, we have a group working on Bayesian networks."
IC: As connectivity becomes more ubiquitous, the policy implications surrounding the Internet are becoming more crucial in terms of economic development, political enlightenment or repression, and cultural cross-awareness. At a general level, do you think policy makers and technologists have been having an adequate dialogue about how their respective work affects each other and, in turn, public access and discourse?
Pearl: The Internet created a new media. It's a new network for news and for opinion — I bet every policy maker's office has a blog watcher — so it's definitely a new outlet for people's frustrations, observations, and information. Unfortunately, there's too much information out there, most of it junk. So, unless we build smarter and smarter search engines, the Internet will eventually render itself useless.
IC: Formal agreements between nations and standards bodies are one thing; journalistic ethics are another. What position do you take regarding the airing of terrorist videos and Web sites?
Pearl: I think there should be a gentleman's agreement among all networks and media outlets not to air any terrorist-produced material. The standard practice should be not to air it, but rather to give it to the authorities for examination. The exception would be to air portions of such videos found appropriate by antiterrorist authorities. But the default assumption should be a total blackout. The only way to block the appetite of the terrorists is to deny them free air time and free bandwidth.
IC: In principle, of course, a journalist is dedicated to free expression.
Pearl: Sure, and that is why I said a gentleman's agreement, not censorship. We have today such gentleman's agreements not to air rape scenes or bomb-making manuals. It's not government-imposed, but is practiced out of public interest. And I don't see why the same situation doesn't apply to terrorist-produced material.
IC: How vital is the Internet to the work of the Daniel Pearl Foundation? Obviously, you have a Web site, and it's your major informational portal, but do you plan to expand the content to include policy discussion forums?
Pearl: I've had the idea of starting a blog, but don't get around to doing it. It would be a good idea to get different types of East-West dialogue in such a blog, which would entice people to communicate ideas and air grievances. The Internet would be a powerful medium for such dialogue. Unfortunately, productive dialogue can develop only in an environment that sustains mutual respect, to avoid ridiculing participants and descending into a shouting match. And such an environment is hard to create without moderation by a professional dialogue-maker, which takes away from its spontaneity.
IC: Do you foresee doing a Web cast, perhaps broadcasting or streaming some of your dialogues with Dr. Ahmed?
Pearl: It's something we've been talking about and it's something we will definitely be considering if we get the Purpose Prize. But the first thing on our mind is to document some of the dialogues and make them available to a wider audience.
IC: How did you and Dr. Ahmed begin your dialogues?
Pearl: I was looking for a "moderate" Muslim to help us reach out to the Muslim community. I asked around and was given Akbar's name at the American University. I rang him at his office, and he was kind enough to set up a meeting. He and I sat down and discussed our views on Jewish-Muslim relations, and he impressed me as a genuinely moderate Muslim … For example, in his book Islam Under Siege [Polity Press, 2003], he empathizes with the way Israelis feel being under siege in a sea of hostile Arab neighbors. Such empathy is virtually absent in the writings of Muslim authors. A few months after our meeting, David Shtulman [executive director, American Jewish Committee of Pittsburgh] called me and said, "Here's an opportunity to make your conversation public, in Pittsburgh, on stage." We ended up in front of 800 to 900 people, talking about facts, dreams, and grievances, what we'd like each community to understand or to do. For example, I expressed surprise that Muslims have not issued a fatwa against bin Laden. The dialogue was discussed and well publicized in the local press. So well, in fact, that other communities have caught on and invited us to continue our discussion in their courts.
IC: Do you consider him a friend at this point?
Pearl: Yes, we are friends. And we demonstrate that friendship on stage. There's also a very important symbolic component of our relationship, the idea that friendship can develop between a Jew and a Muslim, and dialogue can then be built on top of it.
IC: On the whole, are you optimistic or pessimistic about the way the Internet is being used to convey political and philosophical positions on a global scale? Are we taking the best advantage of it that we can?
Pearl: The trend has been positive. Overall, when I look at the progress of the Internet I ask, "How could we think about our lives without it?" Maybe we wouldn't have terrorism today; that's the one negative aspect of the Internet. Yes, there's that possibility that terrorists wouldn't have the ability to communicate and energize cells. I don't know. But you can't stop it, anyhow. You can't stop scientific innovation. And I trust my colleagues to find a technological solution to terrorism as well.
The next scheduled Daniel Pearl for Muslim Jewish Understanding Dialogue will be in Hamilton, Ontario, Canada, on 12 November 2006.