‘Inevitable’ jobs will be more automated, says new AI adviser

Protecting British jobs will be a challenge as artificial intelligence systems become more advanced, the new head of the government’s AI taskforce has told the BBC.

Ian Hogarth said it was “inevitable” that more jobs would become increasingly automated.

The whole world will have to rethink the way in which people work, he added.

“There will be winners or losers on a global basis in terms of where the jobs are as a result of AI,” he said.

This week BBC News is focusing on AI, how the technology affects our lives and what impacts it may have in the near future.

But others believe these developments will also usher in a lot of new human jobs that do not currently exist, just like the rise of the internet did.

He said he was concerned about the potential for AI to cause harm – for example with a wrongful arrest if used in law enforcement, or generating malicious computer code that results in increased cybercrime.

But he was also cautious of not missing the benefits of these technologies.

Notably in healthcare, AI tools are identifying new antibiotics, helping people with brain damage regain movement and being trained to spot early symptoms of diseases.

Mr Hogarth said he once built a tool that could identify breast cancer signs in scans.

The group he will lead has been given an initial £100m to oversee AI safety research.

He would not say how he intends to spend the money but that he thinks he will know if he has succeeded in the job if “the average person in the UK starts to feel a benefit from AI”.

Prime Minister Rishi Sunak has made AI a key priority, and wants the UK to become a global hub for the sector. Someone who knows him put it to me more bluntly: “He is obsessed with it.”

But the UK faces several challenges in positioning itself as a key player in this lucrative and fast-moving area of tech.

Emma McClenaghan and her partner Matt run an AI start-up in Northern Ireland. They have built an AI tool called Wally which generates websites, and have ambitions to turn it into a more general digital assistant.

The company has won awards but they still struggle to access the specialised chips – called GPUs (graphics processing units) – they need to develop their product further.

“I think there is a lack of hardware access for start-ups, and a lack of expertise and lack of funding,” she said.

She said they waited five months for a grant to buy a single GPU – at a time when in the US Elon Musk was reported to have purchased 10,000.

“That’s the difference between us and them because it’s going to take us, you know, four to seven days to train a model and if he’s [able to] do it in minutes, then you know, we’re never going to catch up”.

In an email chat, Ms McClenaghan told me she thinks the best outcome for her firm, Gensys Engine, would be for it to get bought up by a US tech giant – something I hear a lot from UK start-ups.

Re-nosing those ambitions to keep successful firms here in the UK and helping them to grow is another challenge.

Trying to access GPUs might be less of an issue if they were available as part of a national infrastructure, like for example road networks or electricity cables, rather than being hoovered up by those companies which can afford them.

Ian Hogarth thinks this could be the way forward.

“I think we’re going to enter a phase in which nation states start to see their role in directing critical AI infrastructure in a new way,” he said.

“It is going to be a fundamental building block for the next generation of innovation.”

Despite the trials ahead, he is optimistic the UK can still take a seat at the centre of the AI revolution.

“I don’t think we’re too late,” he says.

“I wouldn’t have taken the job if I thought we couldn’t do a lot.”

Follow Zoe Kleinman on Twitter @zsk