Iain Wallace

I've always known all I ever wanted to do was make computers and machines smarter, it was clear to me that it's the technology that will change the world in my lifetime - AI. To that end I had experience of industrial and university R&D prior to my PhD, and my work on autonomous systems, and subsequent MEng work on multi-agent/robot co-ordination motivated the topic which is a more abstract version of the same. Career wise, given I wanted to work at the cutting edge in R&D in industry a PhD set me up well for that.

As a keen fencer and mountain biker Edinburgh and the sports clubs had a lot to offer, with the nearby hills playing a factor in my choice (and the world-class AI research)! Representing the university for both mountain biking and fencing championships and the many trips away with my friends in the clubs are some of my fondest memories of my degree, and the friends I made whilst at Edinburgh. That said, as a researcher, my highlights were probably all the discussions with my officemates in the then-new Informatics Forum - we moved into it during my PhD. Having a great cohort to bounce ideas off and be a sounding board for them was great - many discussions at the whiteboards.

I don't feel I had any uncommon challenges - and they fade with the years since. Managing the workload is one that stands out - and balancing it with demands of two university sports clubs where I held committee positions, competed plus additional training as part of a team sports bursary. You have a lot of freedom in a PhD to work as suits you, but that can also give you the freedom to work a bit too hard... I had a chance to go to Austin, Texas at two days notice to attend SXSW festival which I took... only then did I realise thinking about it that it was nearly a year since I'd had a day with no work. I learnt then to take breaks more often.

I went into my PhD wanting to improve my research skills so I could work in industry at the cutting edge of technology and that's where my career has taken me. Currently, I'm a Technical Director at an AI services company called Faculty. I lead a team that, broadly, delivers AI systems and expertise into companies or government. Frontier AI for the frontlines of the world is the company tagline and it's accurate - we make new things possible for our customers and deliver high impact with AI systems. I've spent my career since my PhD at the frontier - from delivering vision systems for Mars rovers, to CTO of a subsea robotics startup, Chief Engineer Autonomy for an uncrewed air vehicle company and Director of Machine Learning shipping realtime computer vision models for hand-tracking to millions of VR/AR/XR headsets. There's a common thread of trying to do things that no-one knows how to do - because they've not been done ever before, or in a particular application or environment. Tackling novel problems is absolutely core to a PhD and it's a great way to learn that skill.

It's an understanding of what it feels like to learn and develop new knowledge. When you might start not understanding, and then you start to understand but can't question, then you're asking questions and eventually you're coming up with new ideas to solve a problem in a novel space. Knowing that feeling helps me push forward and not be dissuaded from tackling new hard problems.

I was awarded my PhD over 15 years ago now, so I've been asked this a fair bit. Oddly, my normal advice is not to do a PhD! And by that I mean that I often challenge people when asked - most people struggle with a PhD at least at some point. You have to really want to do it, so by challenging and saying "consider if you should" it often inspires people to tell me all the reasons they should and they want to. And that's a great place to start, with strong motivations that set you up for success.

When I started my PhD and was thinking about topics it was obvious statistical ML was the next big thing... so I did something else! This was a good choice, I've had the chance to get deep into ML through my career anyway. A PhD is a chance to be curious, to study something deeply for years - perhaps the only chance you'll get to do this in your life. Be curious!