Ed Lazowska has his gaze firmly focused on the future. He speaks not in terms of the possibilities, but the realities of the coming years. Personalized medicine based on genome sequencing. Web browsers in your brain. Quantum computers. Digital prosthetics. “This stuff is really cool,”
he described.

To hear more of Ed Lazowska's vision, tune in to “Computer Science: Past, Present and Future,” 10 p.m. PST Nov. 21.

Lazowska foresees amazing changes in the coming years, and believes computer science is the ultimate path to this progress.

Lazowska works to prod the process along by inspiring students through his role at the University of Washington as the Bill and Melinda Gates Chair of Computer Science and Engineering, and encouraging forward-thinking research through his position as chair of the Computing Community Consortium (CCC) Council.

In March of this year Lazowska was appointed the inaugural chair of CCC, a consortium of experts drawn from and chosen by the computing research community and funded by the National Science Foundation. He speaks in depth about his work with CCC and shares his vision of the not-so-distant-future in a UWTV program premiering this month: “Computer Science: Past, Present and Future.”

“On the one hand, computer science has accomplished an enormous amount in the past few decades,” Lazowska said. “And on the other hand, the truth is that most people are not really excited by the sorts of research challenges we’re talking about now. Just seems kind of boring, and that’s crazy because there’s a huge amount of cool stuff still to be done.

“So the goal of CCC is to get the field as a whole to be thinking longer range and posing objectives that have more appeal to prospective students, to active students and faculty and to the public at large.”

But before exploring the future, it’s important to consider how far computer science has come.

Lazowska has been witness since the days of mainframes.

“I was an undergraduate and I was flunking out of physics,” he said of his experience as a student at Brown University in the late 1960s and early 1970s. “I took a very early computer science programming course and there was this really charismatic faculty member who sort of got me hooked.”

Ed Lazowska as a Brown University undergraduate in 1971.

One of Lazowska’s first projects was working on WYSIWYG editing systems on Brown’s mainframe.

“We were using a $5 million computer in a several million dollar building as a personal computer between midnight and 8 in the morning and doing stuff that was probably 20 years ahead of the ability of hardware to support it in a reasonable way,” Lazowska described. “And that’s one of the ways that computer science always has made progress, which is profligate use of hardware to simulate the future.”

It was an exciting era to get started in computer science.

“It’s completely remarkable and the progress has been equally astonishing in the hardware and software domains,” Lazowska said. “The algorithms are much better than they were; it’s not just that the computers are faster.

“There’s as much computational power in a Furby as was in the whole Apollo 11 capsule and module. It’s not clear that it’s as good a social good, but we managed to get guys on the surface of the moon with less computing power than is embodied in a Furby.”

“On the software side, it was only in the late 1990s, ‘97 I think, that a legitimate supercomputer won the World Chess Championship. Five years later, you could buy a piece of software that ran on your PC that had the same World Chess Federation rating as this supercomputer. And that’s entirely algorithmic progress, it’s not like desktop computers suddenly became supercomputers in a five-year period.”

These leaps and bounds of past years propel Lazowska to consider what will or should be the advances of the next decades.

“The question is: What are some problems that we can be or should be or are tackling now that have this sort of appeal?” he said. “That are long-range research challenges and that people can relate to?”
Of course Lazowska already has some good ideas.

Among these is meeting information needs around the globe.

“There are 4 billion people in the developing world who have the same information needs that you do and I do, but they have a set of limitations like lousy power, lousy connectivity, no money, low literacy rates that make first-world information solutions not relevant to them,” Lazowska said.

“Another example is your automobile now has hundreds of sensors in it,” he said. “The garage mechanic who used to be a grease monkey with a high school education at best is now somebody who goes to your car, plugs a computer into it, reads out those sensors, finds out what’s going on and fixes it — often by a software fix rather than a hardware fix.

“Now the question is: Why aren’t you the same? Why is it that you don’t know what’s going on in your body? Why aren’t you completely instrumented? Why isn’t your doctor getting a continuous data feed on you? And why, when you go into the doctor, aren’t you presented with a pretty precise description of how your systems are going?

“Over the next 10 years, that’s going to change,” Lazowska said. “You’re going to be completely instrumented. And that’s all about computer science.”

It is upcoming major milestones in medicine that Lazowska seems most eager about.

“We’re going to have everybody’s genome sequenced because the cost is going to drop to 100 bucks over the course of five more years,” Lazowska said. “Before long, we’re going to understand in a lot of detail what your susceptibilities are and what the remedies are. And we’ll do that by correlating millions of individuals’ genomes with millions of individuals’ disease histories. Again, that’s computer science. It’s dealing with massive amounts of data and finding patterns in that data.

“We’re all going to be bionic,” Lazowska said confidently. “And I think that’s absolutely revolutionary. And it’s hard-core computer science, although it’s also mechanical engineering, and chemical engineering and lots of other things.”

Lazowska is also certain about quantum computing.

“It’s a long way from being ready for prime time, at least 10 years,” he cautioned. “We’ve clearly got to think about what we’re going to do to keep things moving forward for the 10 years until quantum computing becomes practical. But it’s revolutionary because it’s an entirely new way of thinking about designing computing devices.

“We have not run out of Moore’s law; that is, transistor density will keep doubling for another 10 years, but we’ve run out of tricks for causing that to yield additional processor speed,” he explained. “The only way we know to get additional processor speed causes the heat dissipation to go up so that the chips melt themselves, basically. So that’s what’s led to this thing called multi-core. What they’re doing is using the transistors to give you two or four, or soon eight or 16 or 32, independent processors on the chip instead of one processor that’s 32 times as fast.

“That’s a problem because most applications are designed in a way that only uses one processor effectively, and so the wheels come off the cart.”

The challenges, Lazowska said, lie in learning how to design and build applications that utilize multi-cores; finding better ways to use transistors and get more speed out of a single processor; and exploring substrates beyond silicon-integrated circuits – such as quantum and DNA computing.
It is these opportunities for discovery that drive Lazowska, now in his 30th year as a UW faculty member. He uses them to motivate his students as well.

“This is a field of unlimited opportunity,” Lazowska said of computer science. “It’s not just unlimited employment opportunity, although there is that: The Bureau of Labor Statistics says that 70 percent of all available jobs in all fields of science and engineering between now and 2014 will be in computer science.

“But it’s also a field of unlimited intellectual opportunity. There’s tons of cool stuff to do. While we realize lots of that cool stuff through programming, lots of the cool stuff isn’t programming. It’s envisioning what you could do. The challenge in computing is to think of new ways to use it and new ways to do it, and somebody has to reduce that to code.”

Tune in to “Computer Science: Past, Present and Future” at 10 p.m. PST Nov. 21.