This month marks three years since ChatGPT was released to the public, a moment widely credited with bringing Artificial Intelligence (AI) to the masses.
In that short time, the chatbot has gone from novelty to necessity for many, helping to draft emails, generate ideas and even make decisions.
But as AI tools become part of daily life, some scientists and technologists are asking whether the convenience comes at a cost.
Globally, AI tools have already reached more than a billion users, according to Microsoft's AI Economy Institute. Around 800 million of those users are estimated to be using ChatGPT, the Large Language Model (LLM) released by OpenAI in November 2022.
These figures point to AI being "one of the fastest adoptions" of a technology in history according to Sean Blanchfield, co-founder and CEO of AI startup Jentic and a member of Ireland’s AI Advisory Council.
"We're so early on the trajectory of this — maybe three years in. If this was the web, we'd be in 1996," Mr Blanchfield said.
In the report published this month, Microsoft, which also markets its own AI products, claimed that the rate of AI adoption has been faster than the internet, the personal computer and the smartphone. Proportionately, people in Ireland are among the most frequent users.
The report estimates that 41.7% of working age adults here are using AI tools daily, a rate surpassed by only the UAE, Singapore and Norway.
But as AI becomes part of everyday life for millions, questions are being asked about what it’s doing to us — not just at work, but in the way we think.
"The promise of it is incredible. And the potential threats are as well," Mr Blanchfield said.
A study released by researchers at Massachusetts Institute of Technology (MIT) in June found that participants showed less brain activity when they relied on ChatGPT to solve problems, suggesting that the convenience of AI assistance "may come at a cognitive cost, diminishing users' inclination to critically evaluate" the AI’s output.
The MIT researchers used brain-imaging scans to measure neural activity as participants solved problems, comparing those who used ChatGPT with those who worked unaided.
READ: Is AI making our brains lazy?
While that study has not yet undergone peer review, researchers in Ireland are also beginning to ask similar questions.
Professor Paul Dockree is a cognitive neuroscientist at the Institute of Neuroscience at Trinity College Dublin (TCD) whose research focuses on how the brain pays attention, remembers, and stays aware. He is concerned the growing use of generative AI may subtly change the way our minds work.
"Traditionally, things we would do, such as writing, problem solving, planning, decision making, ChatGPT is beginning to do these things. So, there's a sense, do those skills potentially weaken over time?" Prof Dockree said.
"If Google Maps shut down tomorrow, we would probably have a few dependency issues in that area. With something like ChatGPT and language, there are possibly bigger implications," he added.
And while researchers are focused on what’s happening inside the brain, others are looking at how automation is reshaping the skills we use every day.
Elaine Burke, journalist and host of the For Tech’s Sake podcast, says concerns about automation’s impact on human skills have long been familiar in aviation.
"There was a study that looked at nearly 10 years' worth of flight data and found that when accidents occurred in 60% of those cases, it was due to a pilot's struggle with manual flying and manual handling of the craft," Ms Burke said.
"As automation is increasingly introduced, they [pilots] are training less in manual control. And that's the thing that you'll need if the technology fails," Ms Burke added.
The concept she’s referring to is known in aviation as the 'Startle Effect’ — the shock or confusion pilots experience when automation suddenly fails, after long periods on autopilot, when they struggle to recall their own trained skills.
Some, including Mr Blanchfield, note that a similar risk may extends far beyond the cockpit, as AI begins to take over some of our everyday decision-making, and we lose touch with the skills we once practised ourselves.
"It's a bit of a paradox. We all have this infinitely knowledgeable personal tutor in your pocket that has access to all of the world's information. But at the same time, a lot of people aren't practising writing," Mr Blanchfield said.
He also believes that while any predictions about the future of AI "will be totally wrong," we no longer have the luxury of waiting to find out. The risks are real, but so is the pace of change, and the focus now, he says, has to be on preparation.
"There was a time early with generative AI where the researchers were really being very careful about releasing it. They were concerned about safety. But then commercial interests took over, and it became a race in to who could release it faster.
"Moving past this, given that now it's here, it's in our lives, [the key] is trying to help people understand what it is, and what it isn't," Mr Blanchfield said.
On the streets of Dublin last Thursday, members of the public that spoke to Prime Time had mixed feelings on the impact the relatively new technology has had on their lives.
Siobhan, a retired schoolteacher, worried about its impact on creativity and people "conceding their ability to think," while Christy, a pastor originally from Sheriff Street in the north inner city said it had improved his life significantly, helping him with life admin that he had previously struggled to keep on top of.
There were differing views from students too – Patrick said it helped with his study and that he used tools like ChatGPT on an almost daily basis. On the other hand, Grace said she felt using AI to study didn’t improve learning and "understanding," of subjects.
Those contrasting feelings mirror a deeper question about how our brains are adjusting to a world where technology is starting to think alongside us.
"There's a potential for interaction between human and AI where there is a collaborative arrangement, an extended cognition," Prof Dockree said.
"Our working memory is quite limited. We're not great at holding information. Maybe there's ways in which we can use AI as a creative partner to challenge our ideas and to present alternatives," Prof Dockree added.
The challenge, he says, isn’t the technology itself, but how deeply we choose to think with it, or let it think for us.
"There may be a risk of a two-tier society where some people are gaining the benefits, whereas other people may be not engaging and are outsourcing and relying on generative AI tools."
A report on this subject from Jack McCarron and producer/director Aaron Heffernan is broadcast on the 11 November edition of Prime Time at 9.35pm on RTÉ One and RTÉ Player.
