Tony Hey on the origins and future of computing

festival_of_ideas17:36 Wednesday 29th October 2014
BBC Radio Cambridgeshire

CHRIS MANN:┬áThe Festival of Ideas at Cambridge University continues tonight with a talk by one of the most senior executives of Microsoft on the future of the computer. Originally from Birmingham, Tony Hey CBE, Professor, is Vice President of Research. His talk will trace the origins and history of computers. He’s written a new book to try and get more young people to understand the origins of computing. He joined me earlier to explain why.
TONY HEY: First of all the inspiration: I used to be a physicist, and I see wonderful popular physics books around. I see almost no popular computer science books, and when I was running a programme in the UK called the e-Science programme, which was trying to get computer scientists to collaborate with scientists, it was quite clear that the scientists, even the scientists at universities, had very little idea what computer science was about. So the idea was to do a popular book on computer science.
CHRIS MANN: And it’s very fitting that your talk tonight is in Cambridge, because this city, this university, has played such a huge part in what you’ve written about.

TONY HEY: Absolutely. So since most of the histories are American, I do feel that they’ve had slightly short shrift in Cambridge, because they built the first really working stored program computers, they developed the principles of sub-routines and libraries. The first computer science PhD at Cambridge was a guy named David Wheeler. He’s known for the Wheeler Jump, because when you do a library, you have to go back to the place in the code, and that was the trick you needed. And that’s now widely used everywhere. So huge contributions in Cambridge.
CHRIS MANN: And of course we can’t forget the name of Alan Turing.
TONY HEY: Yes. I ended up this book being even more impressed by Turing than I’d started. Turing yes, he gave a theoretical basis for computer science, with the fact of what a computable algorithm is all about, how you do it and the Church-Turing conjecture, which enables you to decide this problem is computable, this one is not computable, you can do a lot of analysis. And it all comes down to Turing. But what was even more impressive, nowadays you hear lots of things about machine-learning and stuff like that. While he was at Bletchley Park during the war, he used Bayesian statistical techniques to develop his ways of breaking the Enigma codes. And so he was one of the pioneers of Bayesian statistics, which I did not know until I wrote this book.
CHRIS MANN: Even this early in the history of it all, there are so many claims and counter-claims about who invented this, who developed that, who was important. In your view, who was important, about the internet, and the World Wide Web, which of course are two different things?
TONY HEY: They are indeed. First of all we had the computers with white coats you give your program to and they run it. But actually it was liberating it and giving a computer on everybody’s desk, tablets, iPhones, PCs. And that all started from Bill Gates and Steve Jobs, who made computers usable and available to everybody. One of the unsung heroes of this whole history is a man named J.C.R.Lickleider. He was a professor of psychology at M.I.T. and worked on the early warning system for .. they were worried about Russian bombers attacking, so you had all these things interacting in real time. So he wanted to make computers interactive, and also to connect them. So by connecting them you could do all sorts of things. So he had the idea for the internet. The killer application for the internet was e-mail. Now everybody knows about e-mail, and people in the universities were using it from the ’70s.
CHRIS MANN: Because it linked everyone together.
TONY HEY: Yes and you could send e-mails. So people thought they would share resources. Actually what they did was send e-mails to each other. And that was the key application for the ARPANET and later the internet. But it was Tim Berners-Lee with the World Wide Web who made it very easy to surf the web.
CHRIS MANN: I’ve heard it said that the biggest mistake when WWW was begun was that when you clicked forward, there wasn’t some even small payment back the way, that that would have revolutionised the whole thing. And it would have happened quicker, because there would have been reward.
TONY HEY: Yes. I have great concerns actually about the Web 2.0 philosophy. You can just take anything for nothing and you can mash it up, and that could be the end of journalism, newspapers, artists ..
CHRIS MANN: Well it has been the end of many things.
TONY HEY: The end of many things. And Ted Nelson’s idea was there would be micro-payments. But you see you said it yourself. You had to have a link back to the page that you came from, and that involves the owner of the page having to do something. And Tim Berners-Lee realised the easiest thing was not to do that, and just enable you to just get the things. But I agree. It does make you very concerned about the future, and I think rightly so.
CHRIS MANN: You spoke about Gates and Jobs, and of course you’re one of the very senior people in Microsoft in the world. So you’re helping to plan the future. Where are we going?
TONY HEY: Well there’s great excitement in Microsoft. I’m in Microsoft Research, which is a thousand computer scientists with PhDs looking towards the future. And I would say there’s great excitement. In the early days artificial intelligence, people thought there would be intelligent machines, and that was really over-hyped and never came to pass.
CHRIS MANN: People said that they would take over one day.
TONY HEY: Well we’re getting closer to understanding things. So you know Siri, you can talk into your iPhone and it isn’t perfect, but it can actually do something. And with these new .. it’s called deep neural nets, so it’s a way of doing machine-learning where you practice, and the machine gets better with experience. And now we’re getting close on some tests to what humans can do, the error rates of humans. So some of the classic AI challenges, we’re now getting close to be able to do.
CHRIS MANN: Will the computer take over one day? It’s a basic question people have been asking for a long time. Where do you stand on it?
TONY HEY: I stand with Mr Feynman. Richard Feynman was my hero as a physicist, and his analogy was of a file clerk. It just goes to a filing system and takes out a thing, does what it says, puts the card back, gets the next card. And that’s really all a computer is. It’s a very dumb file clerk, it just does it incredibly fast. And really there isn’t intelligence there.But it can just do things so fast, and so it can do lots of things and simulate intelligence.
CHRIS MANN: Professor Hey, let’s ask about who’s following you. You started in Birmingham, went to Oxford University, now based in California but the world is your computing oyster. With all of the things that you do, delighted to have you back in Cambridge tonight. A lot of the people we’ve talked about are British, most of them are American of course.
CHRIS MANN: Is Britain still able to compete in this? Can we still produce people at the forefront of all of this development?
TONY HEY: I do believe we can. I do find myself correcting. I work on various UK committees as well as us committees, and the UK reports tend to say Britain is world leading. I change that to world competitive, because you only have to see the strength of M.I.T, Stanford, Berkeley, these places, C.M.U. They have huge numbers of very smart people and they have a whole eco-system. Yes, places like Cambridge you have a lot of smart people. We can compete. We have ARM, which is one of the big success stories. But I think one needs to be realistic about what we can do, and I don’t like statements like, we’re world leading. Yes we have people who can be world leading, but one has to be realistic about what one does, and selective.
CHRIS MANN: Final question. Here’s my smart phone here. In a couple of years’ time, five years’ time, ten years’ time, what will it be doing that I can’t even imagine right now?
TONY HEY: OK. So Butler Lampson, who’s a Turing Award winner, works for Microsoft Research based in Boston, talks about three ages of computing. The first age of computing, about thirty years, 1950 to around about 1980, computers were for simulations, actually doing calculations. The second age of computing from about 1980 to 2010 was for communication. That’s Lickleider linking them all up. And you could do all sorts of different things. He regards that we’re entering the age of the third age of computing, from about now for the foreseeable future, which he calls embodiment. And to give you an example, you see these adverts for BMW and Mercedes and so on, where the car brakes before you do and stops an accident. And so it acts on your behalf. And these things will be critical. When I speak into my mobile device, which of course is a Windows phone, and I talk to Cortana not Siri, but actually the calculation is not done on your phone. It goes up to the cloud, and uses super-computing performance in the cloud, and comes down in less than a second, and tells you the answer. And so what you’re carrying around there when you’re connected to the internet is a super-computer. It’s a huge amount of computing power, it’s up in the cloud. You don’t notice that it’s actually not doing the thing there. And that has the potential to transform all sorts of things. If you’re connected to the web, to the internet and to the cloud, you with your smart-phone will be able to do unimaginable things.
CHRIS MANN: Professor Tony Hey, thank you so much for joining me. The name of the book is…
TONY HEY: The computing universe: a journey through a revolution.
CHRIS MANN: And we recommend it to all … geeks?
TONY HEY: It’s for geeky teenagers, but also for the interested lay person. Because I think that’s the normal audience. I’d like to try and excite nerdy teenagers, and get them excited.
CHRIS MANN: And enjoy your visit to Cambridge, and of course your night at the Festival of Ideas. Professor Tony Hey CBE, thank you so much.
CHRIS MANN: Thanks very much indeed.