💬 “Most AI is invisible. It’s embedded behind software systems like social media or in industrial systems. I think there’s a greater public consciousness that AI is ubiquitous, and a growing concern about what this ubiquity will mean for autonomous decision making in and on our lives.”
This is a conversation with John Macintyre
🎙️ John is the Dean of the Faculty of Applied Sciences, and Pro Vice Chancellor at the University of Sunderland. He has a PhD in applied artificial intelligence, focussing on the use of neural networks in predictive maintenance. During the 1990s John established a research centre – the Centre for Adaptive Systems – at the University, which became recognised by the UK government as a Centre of Excellence for applied research in adaptive computing and artificial intelligence. He has successfully supervised PhDs in fields ranging from neural networks, hybrid systems, and bioinformatics through to lean manufacturing, predictive maintenance, and business and maintenance strategies.
🎧 In this episode, John talks about the recent developments in AI that are making the technology broader in capability and wider in scope. He cites the recent GPT-3 release from Open AI as an example. An AI trained on 175 billion data points. John is highly focused on the ethics of AI and so this theme came up many times in our conversation. He talked about the challenges with encoding mortality and the need to slow down sometimes and challenge the work that is going on. John talked about the difficulties we had during the early stages of COVID-19 to make decisions as the data was limited in use. He talked fluently about the types of safeguards and controls that we need in place to use AI ethically, especially as most AI is invisible to the public. We talked about the need for more diversity in AI teams and how this can help with downstream problems like inaccurate facial recognition systems and he drew parallels to the efforts to reduce the impact of man-made climate change.