Two lines on your “claim to fame”
I was part of the Higgs discovery team.
Who are you and how did you become interested in AI?
My name is Ellie Dobson and I became interested in AI due to my background in particle physics, where we used machine learning, alongside many other tools, to discover the Higgs boson.
What is your role at work?
I head up the data science team at Arundo Analytics, a company specialising in data-driven analytics and software for asset-heavy industries.
What are the most important concepts in AI?
It depends on who you ask. To me, it is about finding the best ways to use AI in practice, to drive business value.
Why is this exciting?
AI has been around for a long time, as my dad – whose PhD was in AI 50 years ago – likes to remind me. However, it is only recently that we have begun to acquire enough data to feed the AI algorithms sufficiently so they make useful predictions in a business setting.
What do you think are the most interesting controversies?
I think the key issues are the extent to which AI will replace humans in decision making and what jobs those people will then be doing in the future as a result.
What is your own favourite example of AI?
A good example of AI is how sensor data and maintenance logs can be used to predict failures on complex industrial equipment.
Can you name any other good examples of big data, nationally or internationally?
I’m also intrigued by how sensor data and race information can be used to improve the performance of Formula One cars.
How do you usually explain how it works, in simple terms?
AI is about using historical data to teach an algorithm to link cause and effect from previous examples. That algorithm can then go on to make predictions based on what it has learnt.
Is there anything unique about what we do in AI here in Norway?
We have access to many interesting datasets in the industrial space in Norway. Coupled with its booming tech sector, this makes Norway an exciting place for industrial AI professionals.
Do you have a favourite big data quote?
A bad data scientist will tell you to start with the data. A good one will tell you to start with a question.