Artificial intelligence (AI) could have as big an impact on jobs as the industrial revolution, Sir Patrick Vallance has told MPs.
The former Government chief scientific adviser told the Science, Innovation and Technology Committee that the starting point for AI should be thinking about the benefits it can bring, such as in the field of medicine.
But he warned that the Government needs to be considering the wider impact of AI on society, including how people in some jobs may need to be retrained for other roles.
It comes after the man widely seen as the godfather of AI, Geoffrey Hinton, quit his job at Google and said some of the dangers of AI chatbots were “quite scary”.
With the large models and the potential, how do you determine what's true and what's not? As you can get replication of all sorts of things— Sir Patrick Vallance
Sir Patrick also used his time before MPs to say it would be a “mistake” for the UK not to rejoin the EU science research programme, Horizon, after reports that Prime Minister Rishi Sunak is holding back on such a move.
Asked if he was surprised at how fast AI has moved on in five years, Sir Patrick told MPs: “I think everyone’s been surprised by how much the large generative models have done things that people didn’t expect them to do.
“So that’s what’s intriguing about it – very large datasets, very high computer power, and those models are turning out things that I think even people very close to the field thought ‘actually I wasn’t sure it was going to do that’. It’s done it quicker, it’s done it differently, so yes, that’s been a surprise for everyone, and is a very important development, I think.”
He said the big new AI models would come from big companies “but you need to be able to probe them and understand them, so what’s the core national capability that’s required around that?”.
What happens with these things when they start to do things you really didn't expect and what are the risks associated with that? That may be a slightly longer-term question— Sir Patrick Vallance
He added: “And then I think there are three areas of society that needs to be thought about.
“The first is, with the large models and the potential, how do you determine what’s true and what’s not? As you can get replication of all sorts of things.
“And the second is there will be a big impact on jobs. And that impact could be as big as the industrial revolution was.
“How are we going to think about that, over a slightly longer timeframe?
“And the third, of course, is what happens with these things when they start to do things you really didn’t expect and what are the risks associated with that? That may be a slightly longer-term question.”
In terms of regulation, what do you actually want to regulate? My own view is you don't regulate the advance of technology, but you want to think very carefully about the application— Sir Patrick Vallance
He said the development of AI was a “global issue”, adding: “There’s no point thinking you can protect yourselves against something that you might view as a harmful effect individually, because you won’t. And therefore there does need to be an international position on how to think about this.
“And also, in terms of regulation, what do you actually want to regulate? My own view is you don’t regulate the advance of technology, but you want to think very carefully about the application.”
Asked about views that there should be a pause in the development of AI, Sir Patrick said “unilaterally falling behind” other countries “doesn’t seem to me to be a very sensible approach”.
He said AI “brings a lot of benefits”, adding: “We shouldn’t view this as all risk.
“It’s already doing amazing things in terms of being able to do medical imaging better.
I think it'd be a mistake for us not to rejoin Horizon— Sir Patrick Vallance
“It will make life easier in all sorts of aspects of everyday work, in the legal profession, in all sorts of other areas as well.
“This is going to be incredibly important and beneficial. So that’s the starting point.
“In terms of where it can cause harms and I’ve talked about three areas, I think the most immediate one of those is truth.
“How do you make sure that the output of AI doesn’t distort the perception of truth?
“And that seems to me something that can be looked at, and there would be ways of thinking about marking something as an output of AI where it’s appropriate to do so, so you didn’t end up with people inadvertently thinking that something was something that it wasn’t.
“And I think on the jobs aspect, I think in the industrial revolution, the initial effect was actually a decrease in economic output as people sort of realigned in terms of what jobs were, and then a benefit.
“And I think we need to get ahead of that actually. Which are the jobs, the sectors that will be most affected?
“And what are the plans to retrain and give people their time back to do a job differently, because there will be jobs that can be done by AI, which can either mean lots of people don’t have a job, or it can mean actually lots of people have a job that only humans can do.
“And in the area I know most about in this – in medicine – that could be that you actually get more time with your doctor rather than being pressurised. So that could be a good outcome.”
Sir Patrick also said he believed the UK needed to be part of the EU’s Horizon programme, saying: “I think it’d be a mistake for us not to rejoin Horizon.”
The European funding programme probably took a decade to get going well, so that tells you how long it takes to set up these schemes— Sir Patrick Vallance
Some 15 Nobel laureates have written to the Prime Minister after reports that Mr Sunak is keen on an alternative UK-led research programme – Pioneer – put together by ministers, known as “Plan B”.
Sir Patrick said: “I don’t know where we are in negotiations. I can be very clear, though, that I think Horizon is essential.
“The European funding programme probably took a decade to get going well, so that tells you how long it takes to set up these schemes.
“So the idea that you can instantly set up something equivalent… is, I think, flawed.”
He said there “are some parts of science that need scale, and Horizon brings scale”, adding: “You can’t replicate that scale domestically.”
He said the UK has “been a magnet for that talent that comes through Horizon”, adding : “I think we would be causing ourselves a problem not to continue to be part of it.”
Asked about Pioneer, he said he liked some aspects but “my view is that Horizon is plan A and you’re better off going with plan A”.
