Oxford University Press's
Academic Insights for the Thinking World

Conversations in computing: Q&A with Editor-in-Chief, Professor Steve Furber

The fields of information technology (IT) and artificial intelligence (AI) are ever evolving and with this evolution comes much controversy. There are questions regarding the ethics of AI, the disproportion of women to men in IT, the challenges of cybersecurity, as well as the exciting technologies emerging. We are excited to welcome Professor Steve Furber as the new Editor-in-Chief of The Computer Journal and in an interview led by Justin Richards of BCS, The Chartered Institute of IT, we got some answers to these queries and received a better glimpse into the future of these rapidly changing fields.

Justin: Can you tell us a bit more about your current role?

Steve: At Manchester I am a regular research professor and I’ve served my term as head of department, that’s some time ago now. I lead a group of 40 or 50 staff and students and our general research area covers computer engineering to computer architecture. On the engineering side we’re interested in the design of silicon chips and how you can make the most of the enormous transistor resource that the manufacturing industries can now give us on a chip. On the architecture side we are interested in particular in how we exploit the many core resources that are increasingly available in all computer products today.

Justin: What was the topic of your recent Lovelace Lecture?

Steve: The title is ‘Computers and Brains’ and basically this is a lecture which talks about some of the history of artificial intelligence from some of the early writings of Ada Lovelace herself – 2015 was the 200th anniversary of her birth – through to Alan Turing’s thoughts on AI and then onto the research that I’m leading today, which is building a very large parallel computer for real-time brain modelling applications.

Professor Steve Furber. Image used with permission.

Justin: Can you tell us about your SpiNNaker project?

Steve: SpiNNaker is the massive parallel computer for real time brain modelling and the name is a rather crude compression of spiking your network architecture – it’s not quite an acronym. We’re using a million ARM processors – those are the processors that you find in your mobile phone designed by a British company in Cambridge – in a single machine and with a million ARM cores we can model about 1% of the scale of the network in the human brain. The brain is a very challenging, modelling target, you can think of it as 1% of the human brain, but I sometimes prefer to think of it as ten whole mouse brains. The network we’re using is quite simplified as there’s a lot about brain connectivity that’s still not known, so there’s a lot of guesswork in building any such model.

Justin: You’ve said in the past that accelerating our understanding of brain function would represent a major scientific breakthrough. Can you expand a little bit more on that thought?

Steve: It is clear to anybody who uses a computer that they are incredibly fast and capable at the set of things that they are good at, but they really struggle with things that we humans find simple. Very young babies learn to recognise their mother, whereas programming a computer to recognise an individual human face is possible but extremely hard. My view is that if we understood more about how humans learn to recognise faces and solve similar problems then we’d be much better placed to build computers that could do this easily.

Justin: Where do you see AI processing going in the next five to ten years?

Steve: The big issue with AI is understanding what intelligence is in the first place. I think one of the reasons why we have found true AI so difficult to reproduce in machines is that we’ve not quite worked out how natural intelligence works, hence my interest in going back to look at the brain as the substrate from which human intelligence emerges. If we can understand that better then we might be able to reproduce it more faithfully in our computing machines.

Justin: What about the ethics of AI?

Steve: Ultimately AI will lead to ethical issues. Clearly if machines become sentient then the issue as to whether you can or can’t switch them off becomes an ethical consideration. I think we are a very long way from that at the moment so that isn’t foremost among ethical issues we have to consider. I think there are much more pragmatic engineering issues, for example to do with driverless cars. If a driverless car is involved in a crash whose fault is it, who is responsible? If the crash turns out to be the result of the software bug, if it turns out to be the result of the human interfering with the car, there’s a whole set of issues that will have to be thought through there and they come a long time before the issue of the machine itself having any kind of rights.

Steve’s SpiNNaker project investigates methods of brain modelling. Image used with permission.

Justin: What do you think are the biggest challenges the IT industry faces?

Steve: I think high on the list is the issue of cybersecurity. We are seeing increasing numbers of attacks on IT systems and it’s very technical to work out how to build defences that don’t compromise the performance of the systems too much. So as consumers we install antivirus software on our PCs but sometimes the antivirus software makes the PC almost useless. So there’s a compromise in security, always. Most of us live in houses where the front door will succumb to a few decent kicks, but the bank chooses something more substantial for its vault. Security has to be proportionate to the risk. But I think security is going to loom increasingly large in the IT industry.

Justin: What do you see as the most exciting emerging technologies at the moment?

Steve: The most exciting technologies around the corner I think are the cognitive systems, machines becoming less passive, they don’t just sit waiting for human imports but they actually respond to the environment, interact with it, engage with it and that requires some degree of understanding. I don’t want understanding to be interpreted in too anthropomorphic a way – their understanding may be quite prosaic, it might be at the level of an insect. But an insect has an adequate understanding of its environment for its purposes. That’s how I would expect to see computers developing increasingly in the future.

Justin: What do you think the IT industry as a whole should be doing to improve its image?

Steve: I think the image of the industry is particularly important in the way it comes over in schools and in the choices that pupils make about their future careers. We certainly had a problem recently with the kind of exposure to IT that’s happened in a lot of schools being de-motivating, it has discouraged pupils from computing. I think the changes that are needed to remedy that are now in place and it will take a little while for them to filter through, but of course BCS has played a very active role in seeing those changes through, so hopefully computing will have a better image where it matters most, which is in schools.

Justin: Why do you think that we aren’t seeing so many women going into IT?

Steve: If I knew why women did not find IT so attractive, then I’d do something about it. It’s a major problem that for some reason culturally we think IT and computers are a male preserve and of course if we talk numbers then they are predominately male. It’s a problem that we’ve been worrying about all the time I’ve been in the university and many things have been tried and nothing has really made much difference, so it concerns me hugely but I don’t know what to do about it. I don’t think there is any shortage of female role models, there are plenty of very high-powered women in the computing business. I really don’t understand why the subject is not attractive to girls at school, which is where the problem starts. I welcome any suggestions as to what we can do to remedy this.

Justin: Talking of role models, did you have any of your own?

Steve: My role models were probably not in computing, as I said I came through the mathematics and aerodynamics route at university and was really drawn into computing by what I saw as the new wave of computing based on the microprocessor, which in the late 1970s was a very new approach to building machines. So who do I hold up as a role model? Well, one of the lecturers at the university was John Conway who was always a very inspiring mathematician and it was great fun to listen to his lectures.

Justin: Looking back at your career so far is there anything you would have done differently if you had your time again?

Steve: I don’t think so, there are no decisions in my career path that I particularly regret and I think the advice I give to people is roughly the advice I follow myself, which is to make decisions that keep the maximum number of doors open. So look for opportunities, but when there’s nothing obvious staring you in the face then think about what subject creates the most possibilities in the area you’re interested in. Maximise the number of doors.

The full interview between Justin and Steve was originally published in ITNOW, and may also be viewed on YouTube as a two part recording. Watch Part One and Part Two online.

Featured image credit: Mother board by Magnascan. CC0 Public Domain via Pixabay.

Recent Comments

There are currently no comments.