In a nutshell, tell me about your research?
New scientific experiments, such as the Large Hadron Collider Run 3 and the Square Kilometre Array (SKA), will produce volumes of data larger than anything we’re dealing with today. Extracting information – and impact – from that data will require scientists to use AI to support the analysis process. My research looks at developing domain-specific AI solutions for the analysis of scientific data. These solutions should not only quantify uncertainties and biases correctly, but also preserve the potential for discovery. Most new facilities become famous not for the scientific tasks they were designed to do, but for the new, ground-breaking discoveries they make. Consequently, as we automate more and more of our data analysis in ‘big science’ projects, we need to ensure we are using AI solutions that are designed for discovery – or risk limiting the potential of those facilities.
What do you hope is the impact of your research?
One specific impact that I hope for is to accelerate the extraction of scientific information from large radio astronomy facilities such as the SKA telescope – a huge, multi-location telescope that is being planned in Australia and South Africa. More generally, I hope that my work can increase trust in using AI to support scientific analysis by demonstrating how it can assist analyses that require a precise measurement of uncertainty.
Can you give us a taster of what you'll be discussing at AI UK?
I’ll be talking about the huge volumes of data that we’re expecting to collect from the SKA telescope, and how we can use AI to help prepare for that data and assist us once the telescope is operational. In particular, I’ll focus on the specific AI challenges facing astronomy and how researchers are starting to address them.
Is there a session you’re particularly looking forward to at AI UK?
As well as listening to the other speakers in the ‘AI for science’ sessions, I’m particularly interested in hearing the talks on diversity and inclusion in AI. My own work on understanding bias in scientific data analysis is mirrored by a whole field of study that examines the social biases of AI. As a community, we should understand how our work is reflected in our working environment and vice versa. Highlighting these issues, and their impact, has an important role at AI UK.
If you could have anyone as a mentor who would it be?
Gosh, I have no idea! I’ve been particularly fortunate in my career to have met and worked with lots of inspirational people, so I’ve never felt the need to look for a single mentor.
And finally, when not working what can you be found doing?
In normal times, you’d find me up a hill in the Peak District or the Lake District. These days, you’re more likely to find me in my garden.