Professor Philip Nolan:

I am certainly not an expert on AI and, unfortunately, nor is Professor Gleeson, although he knows more about it than I do. I am quite cautious in this regard because, as researchers, we know one must be careful about commenting outside one's field. I am not going to comment on when or if ChatGPT will become self-aware but there is an important point relating to the previous point on diversity. Much of the research we are doing in artificial intelligence relates to precise applications that would have everybody's support, such as whether AI can support the interpretation of medical imaging. Rather than looking at cervical smear slides or mammograms in the way it was done in the past, AI could scan the slide and draw attention to certain features or pre-process them. There is a significant amount of positive opportunity in AI. A second area in which a significant amount of our research is being carried out is not about the technology but, rather, its societal impact and adoption.

That brings me back to the issue of diversity. In addition to what Dr. Freeman has said, we need to think about the issue of class in research. If the research system is dominated by middle-class people or accessible only to them, the concerns of that class will be represented in the outputs but the concerns of the excluded will not be represented. I am proud that there is a significant amount of work going on in Ireland with regard to privacy, security, reliability and transparency - all the things that will mean the positive and safe benefits of AI accrue but we avoid the pitfalls. I do not know whether Professor Gleeson wishes to comment further. The big challenge is how fast things are moving and the need for us, as researchers, and for committee members, as regulators, to keep pace with that.