Myfanwy Wallwork on data & AI in 2024 and beyond
"Both the technology and regulatory environment have rapidly evolved. What hasn’t changed is human nature - if orgs want to get the benefits that AI promises, they need to understand their people"
Another one of the folk I’ve serendipitously met over coffee, Myfanwy is a fascinating person to talk to. She currently spends her days deep in the practicalities of how organisation can actually implement useful AI strategies and governance processes as AI Governance Lead at the University of Technology Sydney’s Human Technology Institute.
Myfanwy, 2024 has been another busy year for data & AI. What’s one development / milestone / news story that really caught your eye?
The Fifth Quadrant, National AI Centre “Australian Responsible AI Index 2024“ had some great information about takeup of AI, use cases and challenges. The gap between respondents’ understanding of ethical principles and implementation of practice was staggering. About half the organisations who agreed that the principles were important hadn’t done anything about putting those into practice.
You’ve been working in and around data & AI for a while now. Many things have changed! But tell us about something that was true when you started out in this space and is still important today.
Both the technology and regulatory environment have rapidly evolved (call out to the legislation that was already doing the heavy lifting - GDPR, Privacy Act, anti-discrimination Acts, Corporations Act). What hasn’t changed is human nature - if organisations want to get the benefits that AI promises, they need to understand their people. Employees can be invaluable in coming up with new ideas for AI to support overall strategies, as well as pointing out problems before the financial, regulatory and reputational costs of retrofitting come into play. Early engagement is also essential to building the trust that is key to any organisational change; without it, the systems may not be used effectively, or at all. You then end up with a beautifully designed business case, lots of costs and no benefits.
It’s been a heady couple of years with 2024 almost as frothy as 2023. What's one common misconception about AI that you wish would go away?
That Gen AI will instantly make you smarter and faster! … but even if that’s true (or possible), very little thought goes into how that might happen. Starting with a few questions
What do you want to get out of it? Rather than starting with the FOMO position that you MUST use Gen AI, think about what you might be able to do with it. Maybe it could help you summarise developments in Gen AI and the laws, regulations and frameworks that come along with that.
How will you know whether you’re achieving those goals? Start with your baseline; in this case you’re expecting to save time, so estimate how much time you’re spending now. You’ll also need to think about what else you get out of the task, so that you don’t gain one benefit and lose another.
What challenges might you have along the way? You might need training on prompt engineering to get the best results. You’ll also need to find a way to check the results, as the summarisation might miss points that you’re particularly interested in or might be plain wrong. This can also change over time, so you need to plan ahead
The festive season is almost upon us, so many readers will have a bit of extra time to read / learn / reflect. Who do you follow to stay up to date with what’s changing in the world of data & AI?
Leaders in Australia, like yourself, Dr Kobi Leins, Lynn Warneke; the Co-Directors (Professor Sally Cripps, Professor Nick Davis, Professor Edward Santow) and my colleagues at HTI; free alerts from McKinsey, Forbes, LegalTech; podcasts such as Hard Fork.
Leaning into your dystopian side for a moment, what’s your biggest fear for/with/from AI in 2025?
Social divides will widen - for those who can’t access increasingly ubiquitous goods and services which require digital literacy, access to devices and wifi; for those who lose jobs due to the efficiencies promised by AI and aren’t supported with the skills to remain part of the evolving economy.
And now channeling your inner optimist, what’s one thing you hope to see for/with/from AI in 2025?
I’ve had to give up on the robot housekeeper idea due to fears over privacy and surveillance. I love that AI is being used to combat crimes that target our most vulnerable people; by creating a bot that will happily chat all day with a scammer or to detect financial abuse or fraud in bank transactions. The progress that is being made to provide better and more equitable healthcare is also very promising. I hope AI can help address the gap in women’s health, which is driven by lack of data, lack of investment in research and human bias. AI could be used to accelerate research, improve the time to market for new products and address existing data gaps through data augmentation.
You can follow Myfanwy on LinkedIn