Ireland's artificial intelligence ambassador says increased resources will be needed to counter the spread of AI-driven misinformation, as the use of AI tools become more widespread.
Speaking to Prime Time following a special programme on artificial intelligence, Dr Patricia Scanlon said she welcomed increased media and public attention on the sector in recent months.
"We all need as a society to wake up to the fact that we need to take the development of AI seriously. We need to resource and we need to regulate, and then even invest in counter-technologies."
Watch: 'The Future is Now' a Prime Time special programme on AI
Counter-technologies include systems that can be used to detect AI-generated content and prevent or highlight when it is being used for purposes such as fraud, impersonation, fake imagery or propaganda.
She said consideration needs to given to incentivising the development of such technology "to help protect society as opposed to always just building for commercial reasons".
"We have a very healthy AI research ecosystem and I would like to see us invest more in the mitigations as opposed to advancing the technology solely. We need to be able to do this as an altruistic thing. There may not be immediate profit there, but it's for the good of society," she said.
Dr Scanlon has been working in artificial intelligence for 25 years, and was appointed to the role of ambassador in May 2022 with a remit to "lead a national conversation on the role of AI in our lives".
She is the founder of Soapbox Labs, a speech recognition company which has built AI tools aimed at assisting children and educators with language development.
Asked about regulation, she said it would be positive for the industry if properly implemented and resourced with knowledgeable staff.
"People complain a lot about regulation stifling innovation but in actuality it's only the high risk use cases like health, law enforcement, biomedical employment, education - there is very specific use cases - which would be regulated."
She said the pace of change had left many in the industry and policy-making circles scrambling to keep up, but she was optimistic. She believes politicians and officials in Ireland and Europe are considerate of the potential risks and benefits of AI.
In recent months, there have been near-weekly significant developments in the world of AI. Most of the world's top technology companies have released updates to tools, launched new products, or announced funding for AI projects.
It follows the launch of ChatGPT by Microsoft-backed Open AI in November. Its capabilities stunned users and experts, and led to a slew of launches from competing companies.
The pace of development increased concern among experts that risks are not being sufficiently considered.
Earlier this month, Professor Geoffrey Hinton, the so-called Godfather of AI, left his role in Google to speak out about his worry that companies could be developing increasingly powerful systems that could become dangerously "superintelligent", and cause problems for humanity.
"If you'd asked me a few years ago, I would have been dubious about whether it's ever possible in my lifetime anyway, the superintelligence," said Dr Scanlon.
"Now it's kind of a case of, well, we don't know, so let's be honest about it - we don't know."
On Prime Time on Tuesday, Minister of State with responsibility for eGovernment Ossian Smyth said recent developments in AI were "huge" but cautioned against focusing wholly on the associated risks.
"It's as big as the introduction of the web from 30 years ago," he said in response to questions from host Sarah McInerney.
"If you think back to that time, there was a lot of fear about what the internet would bring and part of that was that it would introduce the ability to find how to make nuclear bombs, or it would destroy all our jobs and so on, and so on."
"That's a normal thing that happens when there's a huge technology and it is challenging to us, but there is huge upside."
He highlighted the potential benefits of AI in healthcare, education and other sectors.
"You can see it as a risk or you can see it as an opportunity to transform those professions. So, for example, with architecture - very few buildings in the country are actually designed by architects. This means that it may be possible in future for the majority of buildings to be designed by architects because of the productivity improvement."
He said an assessment of potential jobs impact of AI done two years ago will be updated "because the technology has changed so much".
Asked about differing policies on the use of tools like ChatGPT among government departments, he said more detailed guidelines will be issued in the coming weeks.
