Cambridge Researcher’s Special Christmas Gift
By Rob Knies, Managing Editor, Microsoft Research
In 1825, Michael Faraday, the great British physicist/chemist, had a brilliant idea: Let’s find a way to get children more interested in science. He proceeded to inaugurate the Christmas Lectures, hosted annually by the Royal Institution of Great Britain. Each year, a rigorous selection process is held to determine who will deliver that year’s lectures. This year Chris Bishop, chief research scientist at Microsoft Research Cambridge, has been selected. Having just finished recording of the lectures—which will be televised by British TV’s Channel Five from Dec. 29 to Jan. 2—Bishop recently took a few minutes to discuss the rare honor.
Q: Tell me about the significance of the Christmas Lectures.
Bishop: The lectures go back to 1825, when they were created by Michael Faraday. They’ve been given every year since—with the exception of a small gap during the Second World War—which means this is the 180th year the lectures have been given, so there’s a great historical significance. They are also a great way to communicate the importance of science, particularly to young people. In the U.K., this is the flagship event for scientific outreach to young people, so it’s an amazing opportunity to reach a large audience of young people and get them excited about science and technology.
Q: What do you recall about the Christmas Lectures you watched on television as a child?
Bishop: As I am sure is the case for many scientists, I was a big fan of the Christmas Lectures as a child, and they certainly were a significant factor in influencing my decision to become a scientist. The Christmas Lectures would often be given by very famous scientists, lecturing in their particular field.
The lectures have a very unusual style because they are given to a live audience and are richly filled with demonstrations, props, and live experiments that involve audience participation. An important aspect of the lectures is having children from the audience volunteer to come down and participate in helping to conduct the experiments. It really brings science alive. Today, they are still given in the same lecture theater at the Royal Institution in London where Faraday gave the original lectures.
Q: What was the selection process like?
Bishop: Once the Royal Institution’s selection committee has sifted through its list of nominees and chosen a short list of four or five, those people are invited to do a screen test — essentially a mini-version of one of your lectures. Then the committee takes the screen test and other inputs in order to make its final decision.
For my screen test, I came across something rather interesting. I was reading a book about the history of the Royal Institution, and I discovered a fact I hadn’t previously appreciated. The very first semiconductor was discovered in 1833 by Faraday at the Royal Institution. It was discovered in the basement of the building in which I was doing my screen test. I decided to base my 10-minute lecture around Faraday’s discovery.
I went back to Faraday’s original notes and found a paragraph where he describes this discovery, and in the margin is a sketch of a simple glass apparatus. The material he was working on was silver sulfide, the black material that forms on silver objects when they tarnish. He discovered that when you heat silver sulfide, its electrical conductivity falls, which was exactly the opposite behavior of all other substances known at that time. He described this as “very extraordinary” in his notes, but, of course, he didn’t know what to do with this amazing discovery. It would be another hundred years before people realized that semiconductors can be used to make transistors and microprocessors that would become the foundation of the digital revolution.
So I found a glass blower from the University of Cambridge’s chemistry department and persuaded them to build a replica of this glass apparatus. I bought some silver sulfide and, using modern equipment, did a reconstruction of Faraday’s experiment, having one of the staff of the Royal Institution pretend to be a child in the audience, volunteer to be Faraday, and help do the experiment. That was my screen test. It obviously worked!
Q: How does it feel to be selected as this year’s lecturer, given some of the famous people who have delivered the lectures in the past?
Bishop: It’s hugely exciting. There’s an impressive track record of famous people who have done a superb job of delivering wonderful lectures in the past. You have to remember that the lectures are not delivered by professional television presenters; they’re delivered by active research scientists. It’s a very different environment compared with what we’re used to, and it requires a very different set of skills. But that gives it a different dimension. I have little television experience, as do the majority of scientists who have given these lectures, so presenting in front of a million people is a bit daunting.
On the other hand, I like a good challenge, so I’m thoroughly enjoying it. I feel incredibly honored, and it’s enormous fun. We’re thinking of lots of ways to take ideas from computer science and make them really compelling. What can we do that will really interest and excite the viewers?
Q: What can viewers expect from your lectures?
Bishop: I cover the core ideas of the subject. As the lectures are aimed toward a younger audience, I need to make sure that I really explain the subject and bring it to life for them. It’s a tradition that the lecturer also talks a bit about his or her own research area to give the audience a glimpse of the research frontier, so I touch upon this in the lectures.
There are five 45-minute lectures, and they’re quite broad. They encompass many different aspects of the digital revolution. Let me take you through them:
Breaking the Speed Limit
Lecture 1 is about the microprocessor, one of the most extraordinary pieces of engineering ever. This little chip of silicon, the size of your fingernail, has 400 million components, each of which is 100 times smaller than a bacterium, too small to be seen with an optical microscope. The chip can multiply two 14-digit numbers in less than a billionth of a second. The chips are mass produced, and they cost the price of a second-hand bicycle.
How can you make something that small with 400 million components―and do it so cheaply? I look at how they’re made and how they work. I also discuss some of the problems we’re encountering, why this phenomenal growth in performance is starting to hit some barriers, what we’re doing about that, and what we’re going to do for the next generation of processors. The lecture concludes with a much longer view, looking at some radical new ideas, such as computers based on DNA—futuristic stuff.
Chips with Everything
Lecture 2 is about ubiquitous computing. For every computer that’s produced in the form of a laptop or a desktop, there are a hundred embedded in things like mobile phones, toys, thermostats, and cars. Computers are everywhere, and we’re just at the beginning of a revolution where computers are becoming very cheap and are being embedded into all sorts of things. We don’t interact with them through screens and keyboards. They tend to be dedicated to one particular task. They talk to each other. They talk to the Internet. They share information. They fuse data. They make decisions—often without us even being aware of it.
An important part of this revolution is new display technologies. This lecture looks at some of the new kinds of displays and the way we interact with them in unusual ways, such as touch and gestures.
Ghost in the Machine
While lecture 1 was all about hardware, the third lecture covers software. I try to give people a feel for what software is all about by looking at some of the extraordinary things software can do. I also look at some basic concepts in computer science, such as binary arithmetic or how things are stored on a hard disk, and I use some fun games to explain how they work.
In addition to discussing some of the amazing things we can do with software, I also examine some of the limitations. There’s a tendency to think that computers can solve any computational problem, but there are some computational problems that are so hard the world’s most powerful super-computer would take trillions of years to solve them. Some of these are very practical, real-world, everyday kind of problems. It’s perhaps surprising that computers have these limitations. This lecture ends with a futuristic look at quantum computing, which could overcome some of those limitations. I explain what quantum computing is and how it works. It’s quite an ambitious topic to cover in 10 minutes at the end of the lecture, but I think I’ve got a fun way of explaining it.
Untangling the Web
I’ve devoted a whole lecture to the Web because, after the microprocessor, the Internet is one of the great inventions in the digital world that is having a huge impact. The lecture first looks at search. I can search 15 billion pages in a fraction of a second. How is that possible? How can we transfer information across the Web, and how can we do it secretly? Why is it OK for me to type my credit-card information into a computer on my desk and have it go through hundreds of other computers to a computer in an online shop on the other side of the world? Why can’t somebody on the way just steal this information as it gets passed along?
At the heart of this lecture is one of the most remarkable discoveries in computer science: a way for people to share secret information, even though all of their communications are completely public. It’s very surprising, and I’ve got what I think is a really beautiful demonstration that explains exactly how it works.
Lecture 5 is the closest to my own research interests: Can we build intelligent machines? I look at what intelligence is and why machine intelligence is so hard to achieve. Most of the lecture focuses on one aspect of intelligence: looking at the world around us and recognizing everyday objects such as a cat, a dog, a table, or a chair. It seems simple enough, but doing that is incredibly hard for a computer. Any 3-year-old toddler can outperform a supercomputer at recognizing simple objects.
I look at why that is, and I have lots of nice demonstrations to illustrate the progress we’ve made so far in computer intelligence.
I end the lecture series with a glimpse at some of the big open challenges in computer science. I want to leave the children with a strong sense that we’re just at the beginning of the digital revolution, that most of the really interesting challenges are still ahead, and that there are lots of things for them to contribute— not just by making the technology faster, smaller, and cheaper, but also by tackling some fundamental scientific challenges.
Q: In addition to the televised lectures, what other content will be available?
Bishop: There’s quite a life beyond the lectures themselves. There’s a comprehensive Web site, which is fully interactive. For each lecture, there are a couple of interactive games, which are intended to be fun while also illustrating important aspects of computer science. There are also several downloads per lecture for kids to explore at home or for teachers to use in the classroom. After the lectures are broadcast, the Web site will also feature clips taken from the lectures.
The lectures will be available for streaming from the Royal Institution Web site, and they’ll also be put on a DVD, which will be sent to every secondary school in the U.K.
I’ll be going on a tour of science festivals and giving material from the lectures in different places. I’ll be delivering the lectures again in Japan in July of next year, for broadcast on Japanese television.
Q: Have you received any advice or support from previous lecturers?
Bishop: Absolutely. I’ve made a point of talking to previous lecturers that I have had the chance to meet. They all say pretty much the same thing: It’s lots of fun, it’s enormously demanding, and by the time I’m finished, I’ll be completely worn out, but it’s well worth doing.
David Attenborough makes these beautiful wildlife programmes such as Blue Planet and Planet Earth. He is very well known in the U.K., and he gave the Christmas Lectures in 1973. He talked about the combination of giving five lectures with five different scripts and the fact that you have live demonstrations and children who are delightfully unpredictable because they are picked out at random from the audience and you have no idea how they’re going to respond. He described it as the most challenging thing he’d ever done on television! When he told me that, it added to that sense of being rather daunted by the whole thing.
Q: How has this experience changed you?
Bishop: I’ve taken it pretty seriously and put a huge amount of time and effort into it. One of the things that’s been absolutely wonderful is that Microsoft Research has been enormously generous in terms of allowing me to put aside some of my other responsibilities for several months so I can focus on preparing, rehearsing, and delivering the lectures, as well as creating content for the Web site. That’s been hugely beneficial, and I’m very appreciative.
The fact that I have been able to put so much time and effort into it undoubtedly has helped to contribute to what I think is a really wonderful set of demonstrations. We have about 100 demonstrations in total, many are technically very sophisticated, and quite a few are, as far as I’m aware, completely original.
This is the first time the Christmas lectures have been given on the theme of computer science. In fact, I’ve come across very few examples of lectures of this kind on computer science, so it’s perhaps one of the first occasions where somebody has sat down and thought: “How can I take computer science and explain it to young people, not only using demonstrations running on computers, but perhaps using physical analogies, or everyday examples? How can I take this rather abstract and, perhaps, to some people, rather dry concept of computer science and bring it alive by using, for example, beakers of colored water, or rubber bands and string?”
Q: What do you hope the effects of your lectures will be?
Bishop: There are lots of things I hope will come of it. But the thing I would be most excited about is that some number of young people watching the lectures would say: “Wow, I didn’t realize that computer science was so exciting and so interesting. This is something that I want to contribute to in the future.” Or maybe they get fired up by the physics in the lectures or the engineering or the mathematics. I’d like young people to see this as a wonderful career direction and realize what a fascinating field it is.
Being a scientist is to be paid to do something you love doing. I feel it’s an incredible privilege to earn my living by doing something that’s enormously exciting. As a research scientist, my job is always to be doing something new and to be looking at interesting challenges. If I can convey that message to young people, that to be a scientist means to spend your life doing something that’s absolutely fascinating and incredibly rich and varied and always different and always new, then I’ll feel very pleased.