by Sali Earlsby Sali Earls
Nigel Shadbolt is Professor of Artificial Intelligence in the School of Electronics and Computer Science at Southampton University. He is Director of Interdisciplinary Research within ECS, and Director of the EPSRC Advanced Knowledge Technologies IRC.
Since 1978 he has been carrying out research in Artificial Intelligence (AI) and Cognitive Science, has published some 190 articles on various facets of AI, and has written and co-edited eight books.
In November 2006, Professor Shadbolt was announced as President of the British Computer Society, and he spoke to Sali Earls about his academic work, his ambitions for his discipline, and his objectives for the BCS in this their 50th year.
You're a Professor of Artificial Intelligence - how would you define your subject?
Artificial Intelligence really is an attempt to understand the principles of intelligent or adaptive behaviour - and that's tough. Animals and humans have evolved over billions of years and what we're trying to do is build programs and systems to replicate aspects of how we, and animals, deal with our environment, how we cope, how we solve problems.
It turns out that building a program is a very exacting test of whether you actually understand what's going on or even imagining how it might work, so a lot of the early attempts in AI were trying to find out if we could even get a system to simulate, emulate or behave like a human or animal.
People always imagine AI is like they see in the films - Terminator or HAL from 2001 - some mad brain in a box. That's quite a hard test, because that's not what most of us in AI are about. We may have been inspired into the subject by such thoughts but the idea of building any kind of self conscious program is literally, if ever it's even feasible, some people think hundreds of years away.
But AI, or IA as I sometimes like to think of it - Intelligence Amplification - is where you try and think of assistive technology, and in that way AI has been incredibly helpful. What it's given us is not brains in boxes but a whole range of methods and techniques that are beneficial - for example predictive texting on mobile phones is an idea that's come out of AI, as have some of the concepts that have come out of search engines; and rules based diagnosis for modern car engines.
We've not achieved brains in boxes - , and some people think AI has not delivered - but I think the attempt to understand vision, language, robotics, and more, really have led to insights that we are using all the time.
I think a lot of it is really down to public perception
It really is, yes. There have always been exponents of the subject who have been prone to overclaiming and exaggeration.
You are President of the British Computer Society this year. Why should young researchers or businesses consider joining the BCS?
The good news is that the BCS has nearly 60,000 members, and quite dramatically since 2004, which is when we changed the way in which people come into the Society, we've been adding members at a rate of around 1,000 per month. With the growth since 2004, the average age has come down from 38 to 28. We've got local branches, specialist groups - around 50 of each - and you see growing numbers of people taking exams and using our products.
A young researcher or business practitioner coming in will find networks of like minded individuals, and you can use the network of members in the BCS to good effect.
We've got better at improving our online services. While at university you may not
notice, as libraries provide great access to online journals and books, but once you're out there, maybe working for
an SME it can be really hard to get to this type of material - part of your package on joining the BCS is to have access to thousands of books and journals through book aggregator services that we have.
The e-bulletin that goes round really gives people a sense of what's going on, and what's important in the area. We have a range of practical events including conferences and workshops.
We are trying to build a Society which is much more closely aligned to an individual's developing needs. So we've got continuing professional development products, which kind of enable individuals to get a slice of new qualifications as and when they need it. The CITP qualification is not yet mandatory, but I think as we get to see more professionalism in the subject people will see that, employers will see that as a really desirable kite mark of good practice.
As President, what do you feel your responsibilities are?
I think a lot of it is about momentum. We're lining up initiatives like Professionalism in IT and public engagement - it's one thing to call these things into existence, but to actually execute them and keep them going is something else. It's about keeping the message in front of people.
My theme for this year is trying to improve the public perception of our subject, because unless we can enthral people by what we've achieved in the discipline, and what the challenges are, then you're not going to get the brightest and the best kids into the subject.
Also, trying to work with government, with business organisations, with training establishments, with universities and schools, and with our volunteer base, to keep everyone enthused about what we're trying to achieve.
You made a point there about getting the brightest and best kids involved in Computer Science, but my observation is that the IT curriculum taught in schools does not provide a good grounding in basic computing, favouring something more appropriate to secretarial skills. The perception that many kids have of the subject is one that is often boring or geeky. If you had the opportunity to rewrite the school syllabus from scratch, what would you teach the pupils?
The problem is, if you look at the challenge in schools, I do agree that vocational IT skills are important, and so in some sense IT literacy is a good thing to aspire to. But if that's all we're offering - we do need our workforce to use and interact with software that businesses and organisations use - but that can't be the whole story.
I think it comes back to this issue of inspiration and excitement. Kids at school are particularly susceptible to this, they enjoy their History and their English because they can get into a sense of the excitement of the subject, and I think currently, as you say, much of our curricula is too skills oriented, and they don't see Computing and Computer Science as having been revolutionary in changing the way people think about and approach problems.
So this whole notion of abstraction, or describing processes can capture people's imagination. I think the history and future of the information infrastructure we've built is a great piece of human history as well, and more of that. You can't teach the kids in schools all of the in depth skills to turn them into software engineers equivalent to a second year university student, but you can prepare them to be excited about what they will learn. I don't think we've found that middle ground, and we've got to try and support teachers both with the content we provide them, and the curriculum structure to get down that route.
Because of my background I always look at it from an AI perspective, because I think that's fundamentally an intruiging and fascinating way to come into Computing. The founders of our subject were all intruiged by AI - people like Turing, von Neumann, people like this - wondering if machines could perform in ways that were flexible, intelligent and adaptive.
All those people who founded our subject were excited and pulled in by this, and I think that we could get those messages out much more powerfully, and also it's not just about computers - it's about the information fabric, it's about the web, it's about how information is held, managed and published, and how Computing affects all the other subjects that are being taught at school - we really could pull things together and it could be a very good crossroads for linking subjects together.
I don't really feel that we've got the solution yet.
I think a lot of it comes down to the understanding of the difference between IT and Computing.
I think that as a subject, this is an area where we suffer from people having very strong opinions about what certain words mean, and even if we come up with a good separation and characterisation just finding space in the curriculum I suspect would be difficult.
It must be very frustrating as a university academic when you get a new influx of undergraduates who just aren't prepared for the level of learning required of the subject at university.
Universities also have to face up to the challenge here of what we teach and how we teach it. I think the really alarming thing is that you can see where the problem is in terms of the flow of students coming in - it is frustrating, yes.
Just looking at the UCAS applications for the past few years shows an alarming decline in interest for Computer Science, and you said in a BBC interview that 2009 was a critical year as the economy will decline as more work is outsourced offshore. If young people have lost interest in Computer Science, or just don't get it in the first place what can be done to make them rethink?
I think those numbers really do hold out. We're doing some more work on it with the BCS to really understand those issues, and actually we have more statistics coming in from people like UCAS, so there's no doubt where the trend is going.
The problem in 2009 is that it will hit us then because of where we are in the graduation cycle - people graduating then will be literally where the drop off is most acute.
I think the problem is really complex and there is no one reason, but as I say, I keep coming back to this issue of public perception and engagement. I think we are regarded as somewhat boring, clerical, geeky, that it's just about technology and not about the really deeply interesting aspects that kids are interested in - people, social relationships - and yet everything I see in IT projects is in that area.
I think there's a post dot-com effect. We may be starting to see a change here, but there was certainly a failure to see the opportunities and financial rewards and employment rewards available, but the employment data is looking stronger there.
As I say, it's what they encounter at school, it's not the fault of the teachers, it is in some sense the fault of the curriculum and how students are supported. We've got to bring people with high end Computer Science skills into schools - there are a lack of teachers whose primary qualification is Computer Science.
The IT failures that are touted in the media too often, and our inability to communicate with the media as a subject, I don't think we present stories in a way that is accessible, and hit the angles and issues that journalists are interested in, and this perception that it's not about people.
The challenge really is to dissect this whole issue about perception, because I think that you can change perceptions, it's just that it's going to take a really concerted effort here.
If you think about it, there are few current role models in Computing - no young hero that kids can aspire to follow.
It's interesting, we don't go looking hard for role models, because they won't necessarily be the people who have built the biggest software empires in the world.
Whether it's Page or Brin from Google, or someone else, you want to find people who have done quite remarkable things perhaps on a smaller scale, but that will be really exciting, and here again we miss a trick.
We've got one of the most vibrant games industries in the world and that is staffed by people who are, on average, much younger, much closer to young people's interests. We see this in design as well, and music. We don't really reach out to our skills and strengths in those areas, and there are just some really intruiging people around.
You don't have to be young and funky to be a really good role model and intrigue people - look at other areas. I use people like Robert Winston, Martin Rees (the Astronomer Royal), and David Attenborough, people can be excited by the presentation of an area done right.
Kids want to make a difference, they have a fire in their belly and the issues that matter to them are global warming, poverty, big things and we need to orient the technology we've got around those issues. There are simple things that can be done, but it takes a lot of energy and effort.
What can be done on a local and national level to increase the public understanding of Computer Science?
There is no one solution to this. If you break it down to its components and ask whether the general public is really going to be happy with a diet of chocolate, soaps and reality TV, the evidence shows that this is not the case. The appetite is there - you can see that in the sales of popular science books, and the success of TV that doesn't talk down to people.
I think we need to understand how to engage the media, and how to write for them, and I think that is almost a piece of training that we as a profession and we as individuals should take on.
The technology is a challenge - our
computing infrastructure has become so ubiquitous, and so almost unremarkable, and so almost invisible - it's shrunk in size and increased in power - that it can be
easy to overlook it, and not be rocked back in your socks to say "well, that's absolutely amazing". I think that's something that we need to think about how to get inside the complexity of both the
systems we build, the technologies we rely on and just try and spell that these are the cathedrals of the 21st century. In medieval times, you'd go and stand in a great cathedral and just be awe inspired
by it, knocked back by the incredible human ingenuities. It's hard to see that when it's at nanometres of scale, you're not in the fabrication plans. I was at Bletchley Park looking at the rebuild of the
Colossus code breaking machine. You can be more impressed by that because you can stand in front of these huge
racks of valves and whirring wheels, and you see computing happening in the raw.
So there's a job to do there.
I think the other aspect of this has to do with linking the relevant parts of our subject to the great social and welfare issues of our time. We've also got to work with other professional institutions - we talked about the challenge with schools - I think people like the Royal Academy of Engineering, Royal Society, British Association for the Advancement of Science, Royal Institution, there are lots of other organisations that would put public engagement high on their list of priorities, and we have to find out a way to work together. We can be larger and more effective if we pool our efforts. I think in IT and Computer Science we have these particular challenges which are almost harder than some of the other sciences and technologies face.
The British Computer Society is in its 50th year, and when we look back we can do so with a huge amount of pride at what has been achieved, and the transformation of technology has brought with it a lot of excitement. A lot of it is via schools and the media, targeting groups, the BCS is beginning to launch a programme of public engagement, we are really now taking this on as a big part of our objectives. I think with the growth in members and the security of our position, and the extent to which we can engage with government and business we are in a good position to do that, and we really have to sit down and produce the strategy that executes on it.
You mention linking Computer Science with the great social and welfare issues of our time. Global Warming is the hot button issue of the moment, and statistics show that it takes around 1.8 tonnes of raw materials to make a PC, and there are approximately 300 million computers in landfill. Many large IT companies are adopting a more environmental approach, and "green computing" is now on the political agenda. What do you think small companies, with limited IT budgets, can do to make environmental change a priority?
We looked at this at Christmas time. It is always amazing to see the statistics here - in the UK alone 2 million PCs are trashed. Obviously there are companies that help to recycle, one of which is recycling 75,000 per year to Africa and other places which is great. But for SMEs, how do they get in touch with a company like that, can they even afford to do it? The big corporates can do deals to purchase from really environmentally sound suppliers.
I think for the SME, one of the things it comes down to is that if you want to be green, it's not just down to recycling, it's also about how you use your IP infrastructure. One of the classic ones here is power. We're all probably a little bit guilty of leaving systems on standby, not switching off monitors when we're not using them. So, just trying to get people to be aware that whatever the size of their organisation, there are actual things they can do to reduce emissions.
Then of course procuring, or at least having a look for, suppliers who meet best practice - that information is increasingly available. I think SMEs are also more conscious of the issues surrounding the maintenance and extension of the working life of their machines. The relentless march of Moore's Law means that people are almost obliged to revamp their IT, and the cycle is getting shorter and shorter. But often systems are perfectly functional for substantially longer than people are telling you.
So, it's a mix, and I really do think here that technology is going to help - flat screen monitors are much more environmentally friendly than the old CRT - and hopefully when the European Directive comes in, companies themselves will have to do less.
I think we are going to see a growth in companies that specialise in recycling computer materials, as there is just too much value locked up in it. At the moment we don't even have the specialist sites that places like Germany have to actually recover 100% of the recycled material.
One of your areas of expertise is in the semantic web. As I understand it, this would mean that the internet would be used to share and exchange data and knowledge. Do you think that the much hyped "Web 2.0" fits in with the ethos of the semantic web, and how do you feel about the increase of user generated content?
It's interesting when you see these terms creep into widespread use. I was talking to National Public Radio in the US at the start of the year, and they are already referring to "Web 3.0", which is another way of talking about the semantic web, and increasing amounts of intelligence on the web.
What we're seeing with Web 2.0 is large scale social participation, it's essentially about networks, and an awful lot about data exposure, and also realising that really intelligent web is intelligent not because we've got all these software slots on it, but because we've got lots of eyeballs and brains on the web interacting in ways that had not been possible in the past.
It is really a hybrid system of human and computing elements, and I think in that sense the semantic web is about building a web of data, and not just documents, and what the implications of that are.
The implications are, fundamentally with Web 1.0 when people made their documents available, in many cases they couldn't be at all aware of the surprising ways in which people would use that content or refer to it, so this idea of serendipitous or unforeseen reuse has been an incredibly powerful idea, and the invitation with the semantic web is to do that with much more extensive amounts of data.
For example the information that the Government holds on the places it approves for selling food, or the health statistics of a regional authority. The tendency is for people to put their arms around their data and tell people to clear off and keep away from it, and there's an interesting set of debates here about the rights and entitlements to data that is going to be a growing one. But what we see from various experiences is that you make data available and extraordinary services may result, whether it's a site that gives you the possible sales value of all the houses in a city, or tells you what the current crime statistics are, or something about where the traffic blackspots are, these are really interesting kinds of information service.
With Web 2.0, because it's being driven
by social networks, what we see is that people are starting to feel that we have to be careful about what we do make available, and the terms by which it is made available. But here again social practice
varies. An older generation may be surprised by what a younger generation is prepared to say about themselves on the web. Also the younger generation is much more aware of the ability to create
digital personalities and personas, and actually create different views of themselves.
I think the increase of user generated content can be extremely powerful. So you get a whole group of people working together, and some of our UK agencies are looking at this, and you can suddenly find out a huge amount more about an area than you knew before. On the other hand, you can get a group of people who are coordinating in a certain way that can feel quite oppressive, or risky and you can question whether they should be saying what they are saying.
I think there is a whole issue about how we manage debate and discussions about privacy access, and in a sense our technology has got to mature a bit so we are much more able to express the terms and conditions of reuse of content with the content itself. So it's not just the case that you have a little policy statement on the bottom of your website, actually unless you use the content in a certain way, you can't see the content. I think it's a natural staging post for the kind of web we envisage coming along - the semantic web is going to be a huge amount about the exploitation of very rich and diverse data sets as if they were seamless.
Although a young subject, Computer Science as a discipline should be seen on a par with Physics, Chemistry and Biology, deserving of a representative learned body. As President of the British Computer Society, and a leading academic, what is your vision on the need and future of a single coherent British Learned Society for Computer Science?
I think there is an absolute and clear need. The BCS has always had two aspects: as a professional body to support practitioners, we've got schemes underway for that which is why our ProfIT scheme (Professionalism in IT) is important, qualifications, Continuing Professional Development. This is a long course - medicine took hundreds of years to sort itself out as a profession, we've been at it for fifty, and there's a lot to do.
We've also got to look at the international context as well - we are the British Computer Society, but we've got fifteen overseas branches, and in terms of the standards for the profession you'd like there to be a degree of harmonisation, or at least a degree of understanding of what qualifications mean across the globe and what we mean by professional practice in this area.
Interestingly, the BCS was founded fifty years ago principally by people from business computing who wanted to share the experiences of how they could make these new systems work for them, and they rapidly called in the academic researchers, and then of course it very much became a body representing researchers as well as practitioners.
The learned society issue is really important - we are a professional body, and we are also a learned society, and we want staff and students, and a lot more of the students in our subject to be members of the BCS, and we need to make it a lot easier for them to be as well. We'd like to support them in networks, conferences and lectures; all that stuff needs to be done. We also need to maintain a link between the two and I think that's where the BCS is in quite a healthy position because it's not minded to go one way or the other - there are always different imperatives for the professional as opposed to the educational and research side, but together we're really strong, and I think we have to have a strong representative body.
What technologies do you think society will be depending on in five years time, and beyond?
Clearly the move towards mobile and ubiquitous computing carries on, our low power wireless capabilities are evolving all the time. It's quite a challenge because people could never imagine using all that bandwidth up, could they, and all the airwaves would be free of clutter, but increasingly large amounts of our processing power are going to be working out how to form an ad hoc network in a non intrusive and efficient way.
You're going to see more and more embedded and ubiquitous computing - it will be everywhere from health to transport - there will be "e-everything", and you're going to see increasing amounts of assistive technology because we will need it. We've got an ageing population and IT is not something that you will have had to have been exposed to all your life to still need. As our generations grow and will have been exposed to computing it will be a bit more natural to rely on it anyway, so it will be everything from the internet addressable teapot that is telling you that someone has made themselves a few cups of tea throughout the day and is therefore likely to be mobile and active; through to support for memory impairment using various forms of computer devices.
So I see huge developments in the technology. I also see huge developments in the way society changes its practices to exploit them and they are harder to read sometimes. The way in which new social practices can emerge can be hard to predict - it wasn't obvious to everyone that texting would arrive in the way that it did, and it probably wasn't obvious that video sharing would create a company with the value of YouTube, but these social practices can always be a surprise.
Longer term I think AI, although it won't deliver intelligence in a box, will deliver much more by way of assistive technologies for driving us around, for monitoring us in hospital, even in the domestic context if you look at the home of the future that corporations are increasingly trying to imagine, they see their future as trying to increasingly support all of that activity for leisure, and gaming - gaming is a good example, the most powerful computing network we are going to have very soon now is the sets of connected gaming XBoxes and PlayStations whose processors are more powerful than our laptop processors and whose bandwidth requirements are high - imagine a virtual machine built out of all those gaming elements, and the question is can you do anything other than gaming on them? I can see all sorts of new opportunities arising from these very powerful computing fabrics.
Find out more about Nigel Shadbolt and his work at www.ecs.soton.ac.uk/~nrs. Find out more about the British Computer Society at www.bcs.org.
Send a comment about this article to editor@itwales.com.