'Irresponsible narrative' over AI's harmful effects is misleading, doctors told

“You’ve all been using AI systems for about 20 years, and they haven’t harmed you yet,” he said.

“I don’t think they will either, for a long time to come. Nor will they automate you, so this hype is something we need to be concerned about.” 

Addressing the Irish Medical Organisation conference, he said: “[You hear about] this very irresponsible narrative last year about AI poses an existential risk to humanity, that in some sense, that AI was going to rise up and discover we are all harmful to the planet and wipe us all out,” he said.

It is a grossly misleading and silly statement but there was a political reason for why it was made.

He said while jobs will change, “people will always be required for jobs".

In relation to health, he said: “One of the things that is really difficult for AI systems is balancing complex objective functions.” 

This includes how doctors diagnoses illness, he said.

Professor Patricia Maguire, director of the UCD Institute of Discovery, described how AI can search for new antibiotics — something not considered financially attractive by pharmaceutical companies.

In one study, she said, researchers “took 30bn compounds, used generative AI, and found 13,000 that could work, synthesised them all, and did the lab work.” 

She added: “They found only six compounds out of 30bn turned out to have proper activity [against disease].”

Dr Conor Judge, senior lecturer in applied clinical data analytics at the University of Galway, said AI should be assessed in the same way as new medicines.

“Whatever standard we need for medications, we should apply that same standard for AI interventions in healthcare,” he said. “The key here is prospective randomised clinical trials.”