This was one of the questions posed by Senator Barack Obama this morning to a remarkably high-powered group of panelists at an economic summit, held in the Weigand gymnasium at Carnegie Mellon University this morning. The 2-hour summit focused on innovation and competitiveness, and especially on key questions related to energy and the environment, education, healthcare, national infrastructure, and innovation in science and technology. This was a small, invitation-only event, with about 100 or so people in the audience.

I can’t quite remember all the panelists, but they included: Lael Brainard, vice president of the Brookings Institution; Eli Broad, founder of the Broad Foundation; Geoffrey Canada, president and CEO of Harlem Children’s Zone; Steve Case, chairman and CEO of Revolution Health and former chairman and CEO of America Online; Susan Hockfield, president of the Massachusetts Institute of Technology; Vinod Khosla, co-founder of Sun Microsystems and venture capitalist; Federico Pena, former Secretary of Transportation and former Secretary of Energy; Andy Stern, president of the Service Employees International Union; and G. Richard Waggoner Jr., chairman and CEO of General Motors.

The question about making STEM “sexy” came after almost two hours of discussion on a range of meaty topics, and exposed a general theme of the summit, namely the importance of science, technology, and basic research as the foundation of innovation and economic health. Certainly these are ideas that we know well, and the obvious consensus of the panel on such topics was both important and reassuring to hear. Answers from the panelists ranged from better teacher training, to improved access, to repairing what Obama called our “disastrous immigration policy”.

A few other impressions and highlights:

• The summit was pretty serious and meaty. After CMU President Jared Cohon gave some introductory remarks, Obama spoke very briefly (less than 10 minutes, I think), in order to give the maximum time to the panelists. There was a lot of content, and in fact the topic of science and technology innovation ended up getting cut a bit short.
• Obama was an effective moderator. When a panelist talked in terms of generalities or lofty goals, Obama would press him or her to suggest concrete strategies for federal action. He wasn’t bashful about raising elements of his own policy ideas as strawmen for discussion purposes. Often, panelists would respond with a plea for strong leadership (not just new policy).
• The university wanted to have the name “Carnegie Mellon” appear somewhere on the stage, podium, or backdrop, but this was vetoed by Obama’s summit organizers. In the end, water bottles with the Carnegie Mellon name and tartan plaid were allowed. I noticed that Susan Hockfield, the MIT President, turned her water bottle so that “Carnegie Mellon” faced away from the television cameras. Maybe she did this so that she could gaze at our university’s name. ;-)

The event was, in my view, quite successful. It wasn’t flashy. In fact, at times it seemed subdued and serious. The discussion was earnest, albeit at a very high level of abstraction.

The CMU administration, especially the government affairs office (led by Tim McNulty) and events staff, really must be commended for their role in making this all happen. There were lots of volunteers, too.  Two that stick out because they are from the Computer Science Department are Rob Simmons and Diana Hyde. Carnegie Mellon has become a very, very important (and accommodating) campus for the nation’s policy and thought leaders.

Check out the interview on youtube.

Manuela looks pretty glamorous in her big sunglasses, don’t you think? ;-)

The main attraction was the Turing Award, which was officially presented to Ed Clarke, E. Allen Emerson, and Joseph Sifakis. Ed gave a a touching acceptance speech, citing the importance of his interactions with Emerson (Ed’s graduate student at the time the early research was being done) and Sifakis (who said that he still has some of the original hand-written letters that Ed wrote to him 25 years ago), as well as the 79 (!) graduate students, postdocs, and visiting scientists he has hosted over the years. Ed also explained how his work probably could not have flourished as well at any school other than CMU — something that I’m not sure is really true, but this is still a very nice thing to hear.

Ed’s entire family was in attendance, including his wife, Martha, and his three sons (two PhDs and one MD!). Here is a somewhat grainy photo (taken with no flash on my iPhone) of Martha and Ed, clutching the coveted award, right after receiving it.

It was very nice to see so many good friends and colleagues at the event. So many of them have strong CMU connections, such as Rick and Terri Rashid, Dana and Irena Scott, Alfred Spector, Bud Mishra, Dawn Song, and many others. And, of course, several current faculty members were also there, including Bruce Maggs and Dave Andersen. Randy Bryant, of course, was in attendance. He accepted the Karl V. Karlstrom Outstanding Educator Award for Randy Pausch, giving a very nice acceptance speech “on behalf of two Randys”. Here is a picture of Randy, sitting with Dawn Song and Bruce Maggs.

Of the many great awards given to CMU people (including not only the Turing and Karlstrom awards, but also the ACM Fellow awards to Avrim Blum and Randy Pausch), I was particularly proud of the Honorable Mention in the ACM Doctoral Dissertation awards given to Vincent Conitzer (now at Duke) and Yan Liu (now at IBM Research). Only three Honorable Mentions were given, and so two of the three going to CMU students is really reflects well on our program. (The Doctoral Dissertation Award went to Segey Yakhanin, of MIT.)

Here is a picture of Yan and Vincent, along with fellow student Mugizi Rwebangira:

All in all, it was a great night. Lots of “rubbing elbows” with the movers and shakers in the computing community, and an amazingly strong showing by Carnegie Mellon Computer Science.

Wow!

But now things are cranking up again. It looks like I’ll be on the road for at least part of every week for the rest of the summer. Today I am in San Francisco, for the ACM awards banquet. Avrim Blum and Alyosha Efros will be inducted as ACM Fellows, and of course Ed Clarke will receive the Turing Award. I’ll post a report on the festivities in the next day or so.

The purpose of this article is to comment on the inclusion of Audacity as one of PC World’s 100 Best Products for 2008. Audacity is an open source audio editor, created by Dominic Mazzoni and Roger Dannenberg. It is an elegant, simple, and reliable audio editor. It’s also snappy, making use of a novel data structure that supports constant time editing operations (such as cut, past, undo, redo, etc) — much faster than most commercial editors. Audacity also has a plug-in scripting language called Nyquist. Interestingly, Nyquist is a mostly functional programming language.

Roger’s main research, of course, is in computer music. Recently, his robotic bagpiper, McBlare, was exhibited at the National Academies. An accomplished jazz trumpeter, Roger has developed a range of novel music understanding, improvisation, and authoring systems and algorithms, in some cases going on national performance tours. While Roger has been doing this type of research at Carnegie Mellon for about 20 years, in the past few years I’ve detected a surge of interest in the intersection of computing and the arts. There has always been a tradition of technology and the arts in the School of Computer Science, not only in the work of Roger but also in the Robotics Institute. (For a recent example, see Keepon.) But there are now a large number of very interesting and highly technical artists in the School of Art, School of Music, and School of Architecture. See, for example, the work of people like Golan Levin (Professor of Interactive Media), Noel Zahler (new Head of School of Music), and Mark Gross (Professor of Architecture). At least two new art+technology faculty are being recruited (one in HCI and one in Art). The local community of “hybrids” between computer science and the fine arts is vibrant and growing, albeit still a bit disorganized. I expect this community to gel into a more recognizable entity within the next year or two.

Already, in the education realm, we are seeing a lot of progress. We are launching, jointly with the College of Fine Arts, a new Bachelor of Computer Science and the Arts (BCSA) program. This program, which is quite rigorous, provides a deep blend of computer science with either Drama, Art, or Music. In the fall, Golan Levin will teach an experimental section of 15-100 (intro CS), using Processing (an open-source library for interactive media) as the foundation. Also nearly approved and ready for launch is a pair of programs in Music and Technology, one at the bachelor’s level and another at the master’s level. (A public announcement will be made after the university approves the program.) These new programs, coupled with the already successful Masters in Entertainment Technology program offered by the Entertainment Technology Center, have attracted the notice of a large number of entertainment companies and studios. Watch this space for announcements of these several other new developments, coming soon.

In the mean time, congratulations to Roger and Dominic!

Interesting factoid: Luis tells me that over 750 million distinct users have solved a CAPTCHA. That’s over a tenth of the planet! Wow. No wonder Luis is trying to conquer the world. ;-)

Now, time for me to go back to playing gwap.com…

This year’s hiring season was extremely busy. The CS Department interviewed 10 candidates at the fresh Ph.D. level. In addition, a few other interesting people at more senior levels came by to visit, with varying levels of possible interest in positions at CMU. On top of all of that, the other departments such as ECE, HCI, Machine Learning, Robotics, and Language Technologies, also interviewed many candidates in areas such as interactive media, machine learning, robotics, computer security, human-computer interaction, machine translation, and more. So, in the end, there were dozens of faculty candidates coming through this year.

Whew!

Every candidate invited by the CS Department was great. They spanned many different research areas, such as computational game theory, parallel programming, operating systems, AI, etc. In the end, after careful consideration, the department decided to make just one offer, to Andre Platzer, a finishing Ph.D. student at the University of Oldenberg, in Germany.

We have actually known Andre for the past few months, because he has been a visiting student in Ed Clarke’s research group. Even so, Andre just blew us away during his interview. He works in the area of “hybrid systems“, which is an extremely active (and well-funded) research area in Europe, and is today emerging as “cyber-physical systems” in the USA. Roughly speaking, a hybrid system is one that can be modeled as a combination of discrete systems (e.g., state machines) and continuous systems (e.g., systems of differential equations). Think of a robot that needs to interact with the real world, or an ultra-high speed train that needs to avoid collisions, or a device to be implanted in a person. Specifying, analyzing, and verifying such systems is an area of rapidly growing importance, and requires techniques ranging from logical specifications to model checking to symbolic differentiation.

Andre has developed an awe-inspiring mastery of the mathematics needed to work in this domain and has shown remarkable creativity and cleverness in solving some large-scale, real-world problems. Interestingly, he has also been deeply involved in successful grant proposals worth many millions of dollars. We think Andre will be a star, collaborating with a wide range of people here. And thanks to people like Ed Clarke, we were able to snatch him up without any other US universities even getting a chance to interview him!

Another new faculty member will be Adrien Treuille. Adrien interviewed with us last year but has been spending this past year on a postdoc. He will officially join us this fall, on a faculty appointment made jointly by Robotics and Computer Science.

Adrien is a graphics researcher who studies methods for modeling very high-dimensional nonlinear phenomena, such as large crowds, fluid flows, and human motion. (I’ve poked fun at him, by explaining that his research is all “smoke and mirrors.” ;-) One example of his work that I personally found to be particularly elegant is on model reduction for real-time fluid flow simulation. In this work, Adrien uses classical (and computationally intensive) modeling techniques to obtain models for flows around solid objects. He then computes a “reduced” model that permits near-real-time interaction with the object in the fluid. See his project page for example videos. The potential for both scientific and artistic applications is tremendous. This research was done with his advisor (and former CMU CS Ph.D. student) Zoran Popovic, and was even highlighted on an ESPN telecast of a NASCAR race, to allow viewers to see the “drafting” effect of race cars following each other around a race track.

In addition to his graphics research, Adrien has also been part of a team at UW that has been developing Fold.It, an online game that allows people to compete by folding proteins. I spent a few days playing Fold.It and was impressed with the fit-and-finish of the game and, even more by how addicting it is. Basically, one gets hooked by wanting to spend “just a few more minutes” trying to beat someone else’s score, and of course one also learns something about protein folding.

Finally, it is worth mentioning that one of our Ph.D. students, David Brumley, has received and accepted an offer to join the ECE Department as an Assistant Professor. David is one of the most exciting young researchers in computer network security. He will be working in the CyLab and will undoubtedly collaborate with many faculty in the Computer Science Department.

Andre Platzer and Adrien Treuille are two superb additions to our faculty. While we still have one offer pending (and some of the other departments also have pending offers), I think we can already call this year’s faculty recruiting season in the Computer Science Department a great success.

The commencement activities started (for me) with a meeting with six of our new graduates from the Qatar campus. Wow, what a great group! Bright, energetic, ambitious, and obviously proud of their accomplishments.  All are noteworthy for blazing the trail in our still-new operation in Qatar, though it is perhaps worth singling out Noura El-Moughny, who worked with Bernadine Dias on a senior thesis entitled, “Assistive Computing Technology for Learning to Write Braille“. There were, in fact, several nice student research projects in Qatar, and so I would look forward to Qatar students some day making their way into our 5th-year program and even Ph.D. programs.

On Saturday morning I was in a roundtable meeting with Jeff Bezos, the Amazon.com CEO. Jeff was in town to receive an honorary doctorate, and spent the entire weekend visiting people in both the School of Computer Science and the Tepper School of Business. This was a small affair, with just 6 faculty members. R. Ravi and I were the moderators. Jeff was one of the most fun, engaging, and technical executives I’ve ever met. We discussed many topics, with some of the most interesting having to do with both the technology and the business of the Kindle e-reading device, something that I’ve been enjoying lately. Plans to open up the device and provide an SDK were particularly provocative for me. One thing to wonder about is whether a Kindle-like device might someday be a combined textbook / social-networking device for college students.

At the end of the meeting, I made a small deal with Jeff: I would read my commencement speech notes on my Kindle, in exchange for an autograph. ;-)

(By the way, the Kindle worked great as a mini-teleprompter.) On Saturday evening, I attended the hooding ceremony for the new PhD’s. Norm Augustine, also in town for an honorary doctorate, gave a great keynote, on the now-familiar theme of US competitiveness in science and engineering. He also talked about the many spectacular failures he has encountered — and overcome — throughout his career. These stories, from a person who is among the most successful engineers ever (special consultant to the President; assistant secretary of the Army; chairman of Lockheed Martin; …), were especially inspirational for everyone.

Sunday morning started with brunch and then the main university commencement ceremony. Headlining the program, of course, was the keynote speaker. Al Gore, who described himself as “the former future President”. As we have come to expect, Gore was simultaneously funny and serious, focusing his most important remarks on the problem of global climate change. I don’t think there could have been a more inspirational send-off for our new graduates. As if that weren’t enough, Randy Pausch made a surprise appearance, telling the graduates to stay true to themselves, even in the face of adversity. The graduates applauded wildly.

In the afternoon we held the commencement ceremony for the School of Computer Science. Jeff Bezos, a computer science graduate himself, gave a fantastic keynote in which he revealed the fact that he is an avid Trekkie. (He admitted that, as a kid playing Star Trek with his friends, he always took the role of Computer.) He really charmed everyone and impressed people with his bellowing laugh. Randy Pausch also made yet another surprise appearance, and this time he was given an award from a symposium in Germany that he was unable to attend in person. Randy’s remarks to the SCS graduates were summed up by saying, “Instead of being a computer scientist who happens to be human, be a human who happens to be a computer scientist.” As in the main commencement ceremony, Randy was given a long, standing ovation.

Just think about it. Norm Augustine, Al Gore, Jeff Bezos, and Randy Pausch. Wow, that is an incredible commencement lineup. Congratulations to all of our 2008 graduates! They certainly have earned this and a whole lot more.

• Slashdot
• AP
• BBC
• MSNBC.com
• Pittsburgh Post-Gazette
• Pittsburgh Tribune-Review
• New Scientist

Hope the site holds up under all the load! ;-)

The best Luis quote in these articles? “That is my goal in life, to make (computers) smarter than we are, so that we don’t have to do anything.”

Today, in a truly exciting development, Luis and his group have launched a new web site, gwap.com, sporting four new games plus the now-venerable ESP image-labeling game. I’ve been playing the games for the past couple of days, and I must say that they can be pretty addicting. (The site, being brand new, is still a bit shaky, but is pretty slick and works reasonably well already.)

Each game has the same general framework. Consider ESP. You start it up and it finds someone else on the Internet who also wants to play. You and your partner are shown the same image, and now the goal is to guess what word your partner is thinking, based on viewing the image. The faster you guess a word that your partner also guesses about you, the faster you score a point and are given the next image. The goal is to score as many points as possible in three minutes. Another game, called Squigl, is similar. In Squigl, you and your partner are shown the same image and asked to trace one of the objects or features in the image. (For example, “trace the dog in this picture”.) The more similar you and your partner’s traces are, the more points you get. Other games test music categorization, word association, and beauty.

The utterly brilliant thing about these and the other games is that they utilize the special abilities of people in order to make computers smarter. In ESP, for example, the matching words are used to label the images, for doing things like Google image search. For Squigl, it is clear that one use is highly sensible image segmentation (and for the words that are provided by ESP!). One of those truly rare and amazingly elegant ideas.

For the next year or so, Luis and his group are hoping to learn a lot more about these sorts of games, as part of a larger research agenda to understand “human computation”. With greater understanding, it might be possible to understand how to harness thousands or millions of people to solve all sorts of important problems. As this understanding solidifies, we hope to establish a major research focus here at Carnegie Mellon and other places.

Finally, it is worth mentioning how much fun Luis and his group seem to be having. They are getting lots of press coverage. (See the BBC coverage and today’s Pittsburgh Post-Gazette article, for example.) A list of the team members can be found on the gwap.com news page. One nice glimpse into the group is given by this short video. It explains, without the use of any words, the idea of “games with a purpose”.