I focus on investments and advisory in AI including technical due diligence and helping clients build real-world AI applications. I also build machine learning applications for Alpha Intelligence Capital and teach data science courses at the University of Hong Kong.
(Computer Vision PhD, 2008)
Tell us about your PhD at Queen Mary. What thesis did you explore?
When I was an undergraduate student in Hong Kong back in 1999, the most interesting courses were two courses in AI, “Computer Vision” and “Neural Networks”. I decided to do my PhD in computer vision after my master’s by research degree in astrophysics. My supervisor, Prof. Shaogang Gong and I looked into object tracking which is an important problem in computer vision. I have always been very interested in the theoretical side of things in general. Machine learning has an important set of fundamental techniques for computer vision which are both theoretical and practical. The research environment is excellent at Queen Mary and this was a key driving force for me choosing to study here. I also got a college studentship which allowed me to study in the UK for my PhD. I then moved to Austria for my postdoc to work with Peter Auer after finishing my PhD at Queen Mary to explore my interest in machine learning further.
How has your PhD and your time at Queen Mary helped shape your career?
My supervisor guided me throughout my PhD and supported me to do two internships in the US and France to get industrial experience during my studies. This has paid forth into my career by giving me invaluable work experience to complement my academic experience. I also met a lot of interesting people at Queen Mary which helped me develop my interpersonal skills. Many of these people have turned out to be very successful which is very inspiring. Back in 2003 when I started my PhD, much fewer people outside of the research community talked about AI but my PhD allowed me to pursue what I am interested in and has allowed me to progress in the industry I now work in.
What did you enjoy about your time at Queen Mary?
Although the population density in London is very high, there is a nice campus at Queen Mary and the people are very nice too - there is a real sense of community. In general, I never experienced any issues when I was in the UK and I really enjoyed my student life living on campus. The people at the accommodation office were always very helpful if I needed their assistance. I still owe them a big thank you!
With recent funding and a lot of organisations getting involved, many proposed AI applications discussed mostly within the research community before are making their way into the real world. It is great to see this actually happening.
What do you do in your day-to-day role(s)?
I focus on investments and advisory in AI including technical due diligence and helping clients build real-world AI applications. I also build machine learning applications for Alpha Intelligence Capital which is a venture capital fund, and I teach data science courses at the University of Hong Kong as a visiting assistant professor. I’ve learnt a lot on what I am passionate about over the years since I graduated from Queen Mary at the end of 2007 and this gives me real satisfaction.
After obtaining my PhD from Queen Mary, I got involved in an EU-funded project on image retrieval using eye tracking called PinView. I came back to Hong Kong and became a quantitative analyst to apply machine learning to algorithmic trading at an AI company called Sentient Technologies based in Hong Kong and San Francisco. I also worked on a number of interesting machine learning projects with applications to drug discovery, smart transportation, materials science and astronomy when I started teaching as an assistant professor in 2015.
What are some recent AI innovations that you find particularly interesting?
With much more funding going into AI, there have been quite a lot of interesting recent models like GPT-3, generative models for videos and machine learning for neuroprosthetics. I dreamt about very practical AI applications when I was an undergraduate student twenty years ago. With recent funding and a lot of organisations getting involved, many proposed AI applications discussed mostly within the research community before are making their way into the real world. It is great to see this actually happening.
Where do you see the AI industry in the next 5-10 years?
In the next 5 to 10 years, if funding continues to come in, there will be a lot of interesting practical applications and solutions and we will be propelled even more so into the future in terms of technological advancements. However, just like the Internet after the 90s, no one knows when there will be a market correction and some companies can certainly be overvalued. But I think useful AI technologies creating real value will survive, even during a recession, just like technologies with the Internet after the dot-com bubble.
What do you think the long-term future of AI will look like and how will it apply to our everyday lives?
Augmented intelligence will let us get smarter. AI is not there to replace us. It will help us to make better decisions. Autopilot systems can be made smarter with improved control using state-of-the-art reinforcement learning and computer vision which will help pilots fly better and more safely for example. In fact, Airbus demonstrated fully automatic vision-based take-off last year!
Augmented intelligence will let us get smarter. AI is not there to replace us. It will help us to make better decisions.
Throughout your life so far, you have studied and worked in multiple countries, how has the approach to and interest in Computer Science and AI differed around the world?
Culturally people are all very different. Some people are better at certain things. There are very creative and hardworking people. There are theorists and experimentalists. This means that some countries are more advanced in terms of AI than others and some countries understand the full potential and practical applications compared to other countries. However, I believe in collaborative knowledge with different people from all sorts of backgrounds to make the world a better place and to use AI for the greater good.
Democratizing AI involves a great deal of engineering efforts. Some countries in East Asia are catching up very fast when it comes to engineering and hard work. However, AI breakthroughs require both theoretical and practical insights and at the moment, the best AI theories are developed in Western countries. Countries without too many regulations on data privacy can actually build a lot of AI applications very quickly as quality data is one of the most important elements. Hopefully in the near future, AI developments will be more easily accessible and possible globally.
What would you say to a student considering studying Computer Science and AI?
Do it. There are a lot of opportunities career wise once you graduate. People get hired into data science roles, AI consulting, finance, insurance companies and all kinds of AI startups (with autonomous vehicles for example). More possibilities will open up in the future too as people and companies global wake up to the power of AI.
What advice would you give to students and graduates wanting to pursue a career in Computer Science and AI?
Passion is what makes someone a persistent individual to achieve his/her goal. If you are genuinely interested in something, you’ll have fun solving the problems along the journey to success even if you don’t make it at the end. Use your passion to drive your career and to find a particular niche you can contribute to.
This profile was conducted by Alumni Engagement Officer, Nicole Brownfield. If you would like to get in touch with Alex or engage him in your work, please contact Nicole at n.brownfield@qmul.ac.uk.