Visitors at Beijing’s Tsinghua University for Microsoft Research Asia’s Computing in the 21st Century conference on Oct. 24 could have been excused for mistaking the event as a TV extravaganza. Eight huge photos of the featured speakers decorated the entrance to the school’s 4,000-seat gymnasium. A collection of video screens provided close-ups of the presenters to attending students and faculty. A multicolored, arching lattice provided a focal point for a flower-festooned stage. Electricity was in the air.

Welcome to Computer-Science Idol!

Kurt Akeley, Microsoft Research Asia general manager and principal researcher, acted as master of ceremonies for the eighth annual installment of the conference, which brought together a clutch of computer science’s leading lights, from Microsoft Research and from academia, to discuss the show’s theme, The Challenge of Quality Software.

The Beijing event represented the first half of a two-day event, which resumed on Oct. 27 in Taipei, Taiwan, in the Grand Ballroom of the venerable, traditionally ornate Grand Hotel. While the venue’s wood-paneled interior and gold-fringed red-velvet curtains may have exchanged glamour for glitz, the impression of grandeur was no less striking.

And appropriately so. The conference featured no fewer than three winners of the A.M. Turing Award, considered computer science’s most prestigious accolade. Jim Gray, a San Francisco-based Microsoft technical fellow who won the honor in 1998, was joined onstage in both Beijing and Taipei by Andrew Chi-Chih Yao, a professor with the Center for Advanced Strategy at Tsinghua University and winner of the 2000 Turing, and Niklaus Wirth, professor emeritus of Computer Science at the Swiss Federal Institute of Technology and 1984 Turing Award recipient.

The procession of industry and academic star power didn’t end there. Also featured were Rick Rashid, senior vice president of Microsoft Research, and Harry Shum, Microsoft distinguished engineer and managing director of Microsoft Research Asia. The Beijing festivities also featured Eric Grimson, department head of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology; Pat Hanrahan, Canon USA professor of Computer Science at Stanford University; and Ed Lazowska, Bill and Melinda Gates chair of Computer Science and Engineering at the University of Washington. In Taipei, Hsiao-Wuen Hon, general manager and principal researcher of Microsoft Research Asia, joined the splendid array of IT visionaries.

With quality software as the unifying concept, each participant offered a unique perspective on the theme. The procession of brilliant minds and their particular musings left a combined assemblage of 5,000 dazzled by the intellectual vigor on display and enthused about the potential of computer science to continue to transform the lives of billions in ways scarcely imaginable a generation earlier.

On both days, the audience was greeted by Akeley and representatives of co-organizing institutions—in Beijing, Jia-Guang Sun, vice director of the National Natural Science Foundation of China, and Kejun Kang, vice president of Tsinghua University; in Taipei, Wen-Tsuen Chen, president of National Tsing Hua University, Wei Hwang, acting president of National Chiao Tung University, and Tai-Jen Chen, vice president of academic affairs for National Taiwan University.

Then, in Beijing and Taipei, Rashid took the stage to discuss Microsoft Research’s 15 years and what the organization has accomplished over that span. He outlined progress made in software engineering and systems design, ubiquitous input/output systems, the graphics revolution, and the future of storage.

“Looking back, it’s been a pretty amazing 15 years,” Rashid said. “I think we’ve done a tremendous amount to change the nature of software, of graphics, of computer vision, of speech recognition, and many other fields. Looking forward, the future is incredibly bright.”

He also urged attendees to read Towards 2020 Science, a 2005 report produced under the oversight of Microsoft Research Cambridge that presents a vision for the evolution, the challenges, and the potential of computer science and computing in scientific research over the next 15 years.

“You can see some of the key things about computing that are going to drive the future of technology,” he said. “It’s an incredible report, and it gives you a sense of how computing technology and technology in our field are going to change the world for the better.”

Yao shared some of the fruits of his research with his talk, entitled A Modern Theory of Trust-but-Verify, addressing what he called “one of the most fundamental questions asked in science.”

“Two thousand years ago,” he said, “Plato asked, ‘What is truth?’ That, perhaps, is the oldest and the original question for scientists. Throughout the years, what is truth and how truth can be established has been the pursuit by mathematicians, logicians, and computer scientists.”

He went on to describe a theory of proof by interactions, in which a statistically relevant set of random checks can serve to prove a transfer of knowledge virtually unable to prove by conclusive analysis. This, Yao said, has particular relevance to software development and the new environment of network computation ushered in by the Internet.

“We have to adopt an evolving and interdisciplinary attitude,” he concluded, “to solving various questions.”

Fellow Turing recipient Gray took a more broad approach in his discussion of eScience.

“Computers,” he observed, are revolutionizing almost every aspect of the way we do things.”

He defined eScience as the synthesis of IT and science, and noted that science is being codified, that scientists must cope with an increasing deluge of data, and that the scientific community is rapidly changing in response to online opportunities.

“eScience is largely an attempt by the computer-science community to help the other sciences,” Gray said, “to manage the information that they are collecting, analyze it, organize it, and visualize it.”

He traced the evolution of scientific paradigms from the empirical explorations of antiquity, to the theoretical advancements of the last few hundred years, to the computational era beginning in recent decades, to the data-exploration stage currently under way. After making suggestions on how computer scientists can engage successfully with those from other disciplines, he examined a few eScience projects of note, and ended with an impassioned argument for freeing online scientific literature from the cost-dictated bounds of print publication processes.

“eScience is about reconceptualizing the way we do science,” Gray said. “It’s about combining IT technology with the various sciences to make science more productive, which is probably one of the most effective things we can do.”

The third of the Turing Award winners to speak, Wirth, cast his gaze backward, delivering a talk on the history of software engineering. He traced the discipline from its genesis in the late 1950s and the early 1960s, from the era of mainframes and punch cards to the adoption of tools and formal notations, and the development of various programming languages—including his own creation, Pascal. And while citing some of the seminal papers in the field—from the likes of E.W. Dijkstra and Tony Hoare, currently a principal researcher at Microsoft Research Cambridge—and the burgeoning popularity of “toy-like” microcomputers of the mid-1970s, Wirth also noted the increasing complexity of computing and the challenges that presented to software engineers, challenges that linger today.

“The increase of computing power,” he stated, “was itself the reason for the terrifying growth of complexity.” The result, he argued, was duplication, inefficiency, and the wasteful use of cheap resources.

And the solution?

“Programming is not only finding the perfect code for a given problem of achieving optimal performance and economy for our computers,” Wirth said. “Ideally, the programmer should achieve satisfaction in his work through a program’s beauty and elegance, two properties that are difficult to quantify. A good piece of engineering is also a piece of art.”

Shum brought a bit of levity to his talk on Research 2.0, which focused on the opportunities for using the Web as a research platform, thereby offering the possibility of deployment-driven research.

The Web, he noted, constitutes a huge database, a conduit for social networking, and a computing platform, connecting data, people, and intelligence, respectively. In this scenario, data leads to knowledge, which leads to value, through a process of finding, aggregating, understanding and sharing. The very nature of the Internet enables users to serve as co-developers in a process of delivering software and using resultant, nearly instantaneous feedback to provide rapid enhancements, thereby shortening the innovation cycle and reducing the time to publication or to market.

But it was with a couple of demos that he was able to bring down the house, with a bit of collegial assistance. The first, involving Microsoft Research Asia’s Chinese Couplet Generator, takes a user’s initial sentence and extrapolates a companion sentence to complete the couplet, an important linguistic device often deployed in Chinese art and culture. Demonstrated in Taipei by Hon, the demo elicited waves of enthusiastic applause from an appreciative audience.

Following Hon, Shum then provided a preview of technology called Video IM for Fun, in which a user can disguise his or her video image to varying degrees, such as increasing the sizes of facial features or adding features such as sunglasses, a mustache or even a mask. The humorous images of the lab director looking big-nosed or bug-eyed convulsed the crowd with laughter.

During the Beijing event, academics Grimson, Hanrahan, and Lazowska took turns wowing their audience with their particular discussions. The former supplied a fascinating examination of the use of computer science in medicine, called Image-Guided Surgical Systems, in which computer-vision technology enables brain surgeons to see through the surfaces of the brain to identify the position and the shape of internal structures, thereby improving the precision of surgical techniques.

“For neuroscientists,” he said, “this is a very exciting finding, because it identifies the specific parts of the brain that appear to be implicated by the disease. It tells them things about what parts of the structure are affected. This is a place where computer vision, computer science, and machine learning are being used to change the way medicine is practiced.”

Hanrahan focused on computer graphics and stated that representing images procedurally enables the creation of elaborate physical simulations and representations of abstract concepts. The procedural approach, he said, enables building of tools, systems, and hardware that are becoming integral components of computing.

While Wirth had the final say in Taipei, in Beijing, Lazowska delivered the concluding lecture, which he entitled Computer Science: Past, Present, and Future. Putting his emphasis on the next 10 years, he spoke on a dizzying array of topics: rearchitecting the Internet for increased security and reliability, “flattening” the world via distance learning, the potential for neural engineering and computer-aided driving, personal-memory storage, parallelism in computing, personalized health monitoring, quantum computing, computational and synthetic biology, virtual conferencing … The list was long, the vision mesmerizing.

“I think, Lazowska summarized, “there’s never been a more exciting time for computer science.”

For audiences rapt amid sumptuous settings in China and in Taiwan, that sentiment certainly rang true.