Dr Fumiya Iida, University Lecturer in Mechatronics in the Machine Intelligence Laboratory and a Fellow of Corpus Christi College, wrote the article below on artificial intelligence for CAM Magazine.
I’m still wondering how to strike a balance between ensuring the detailed accuracy of my statements as a scientist and the important obligation to report research progress to taxpayers and stakeholders.Fumiya Iida
A few weeks prior to the start of the academic year, the BBC reported on my team’s research into how machines might replicate the autonomy and adaptability of biological systems, particularly at the level of evolutionary process. The project’s technical progress was reported as it had often been done before, but this time the result was something of a media hurricane.
On arriving at the BBC on the day of the broadcast, a producer told me: “We love you at full speed today, then forget you tomorrow.” I didn’t understand what he meant until he showed me the schedule for the day: 14 interviews for various radio stations starting at 8.13am and then every seven or eight minutes thereafter. This was just the beginning. At 10.15am, I was dragged to see the make-up staff before a 10-minute TV interview; meanwhile, my mailbox and mobile phone beeped continuously with requests for interviews that afternoon.
Of course it’s important to communicate what we do to the wider world. But looking back at the coverage now, I realise that the momentum for the story largely originated in people’s fear of ‘the rise of machines’. In some ways, it’s not surprising: ‘Machines becoming creative through evolution-like processes’ is a pretty eye-catching headline. The problem is, of course, that it disregards entirely the current maturity of technology – the main concern, from a researcher’s viewpoint. I’m still wondering how to strike a balance between ensuring the detailed accuracy of my statements as a scientist and the important obligation to report research progress to taxpayers and stakeholders.
After such an exciting start, you might think that the rest of my term might be quieter – but it has not worked out like that. In what was effectively my first ‘working’ term since I joined Cambridge last year, I taught a new lecture series, supervised a new module, was affiliated to a College for the first time, and acquired six new research students and some new administrative tasks to boot. To manage all these new tasks, my entire calendar was optimised to the minute (and I’m very grateful to my family for their patience).
But despite the sometimes overwhelming challenges, I have been extremely excited to embrace the work because of the ambition I have had in my mind for a while – to create a robotics culture and curriculum for the first time in the long and illustrious history of Cambridge engineering.
Unlike many engineering subjects, robotics is a field of integration and interdisciplinarity. To build robots we need knowledge from electrical, mechanical, material and computer engineering, as well as from maths, physics, chemistry and biology. Robotics needs all of these disciplines, but at the same time, if a robot is dissected into a smaller piece for the sake of simplicity of investigation, the research is valid – but it is no longer robotics. In other words, robotics is a study of something that is more than the sum of its components. It is in many ways philosophical and also often subjective. There is no simple way to structure a roboticists’ community for these reasons, but the culture needs to be built up on top of many creative minds with a rich diversity of backgrounds.
Fortunately, I am not alone in this ambitious vision – there are many keen roboticists in Cambridge. As this year goes on, I am thinking more and more about ways to bring the robotics enthusiasts together to create a fertile robotics community. And who knows – despite the negative spin, the blanket coverage of robotics one day last September might spark the curiosity of a few more.