Welcome to Jimmy's Jobs of the Future. It is impossible to escape AI at the moment.

It has been the buzz topic on everyone's lips for 2023.

But today we speak to Michael Osborne, a professor from the University of Oxford, who has

specialised in AI for the best part of two decades. We speak to him about where he sees

the future going and how he's using AI to help him in his job as a professor.

Now he wrote a landmark paper 10 years ago saying that AI would lead to 47% of job

production. How close were you in that prediction?

Right. Well, firstly, thanks for your identifying our paper as being a landmark.

But I have to correct you in that what we were talking about was not job replacement.

Instead, our topic was the scope of technology to perform tasks that today are done by

human workers. And this is quite a subtle point, but it's an important one. So,

forgive me for kind of diving into it. The first thing to say is that jobs always change,

kind of tasks that we're doing in our jobs today are already quite different from the tasks

we're doing 10 years ago. And if you look over even longer horizons, for instance,

what doctors do today is completely different to what doctors were doing 100 years ago.

The point is when we were saying that 47% of work might be susceptible to automation,

we weren't saying that 47% of workers would become unemployed. Because work is always changing,

and at least in some occupations, we'd expect technology to boost demand to actually result

in there being more jobs for those workers. And software engineers might be an example where,

despite technology becoming more and more capable of doing what software engineers have historically

done, there might yet be more software engineers around as a result of the technology stimulating

demand for software. But the media undoubtedly gets carried away with predictions and so on.

...