AI will bring a revolution in the jobs market in the coming years a government paper has warned.
The automation of roles could lead to mass unemployment but it could be offset by the creation of new jobs.
The government is due to host the world's first AI Safety Summit next month.
The government has warned the coming artificial intelligence revolution could upend the British job market, with several industries facing the prospect of being almost completely automated.
Ahead of the AI Safety Summit - the first major gathering of its kind due to open on 1 November - the government published a discussion document detailing the risks AI could pose to the UK and wider society.
The document highlights one of the biggest changes will be in the labour market, making comparisons to previous major innovations in the 1960s.
It points out 60% of the jobs that exist today did not exist in 1940.
The document also points out AI replacing jobs may not lead to mass unemployment as new roles may be created at the same time.
It says mass unemployment was avoided in the 1980s with the introduction of automation due to new roles being created.
What roles are under threat?
The paper included a graph published by Goldman Sachs in March 2023 that looked at which sectors could be automated the most by AI.
It shows that IT, financial and legal jobs are most at threat of being replaced by AI while education, agriculture and construction were the least.
Most at risk was the legal profession which could see 40% of its roles replaced by AI and the other 60% be complemented by it, likely also reducing jobs.
Whereas building and grounds maintenance had only 10% of their roles that could be supported by AI.
(Scroll to the bottom of the article for a close-up of the chart above)
In all, the Goldman Sachs research paper said up to two-thirds of current jobs "are exposed to some degree of AI automation, and that generative AI could substitute up to one-fourth of current work."
But the bank also says this could be offset by job replacement in new industries and points out it could add significant growth to the global economy.
How else could AI change the job market?
The government's paper also warns of the dangers of imperfect AI that makes poor decisions.
It warns if biases are not ironed out of an AI those biases could be reflected in its decision-making.
The paper warns this could have an impact on "job recruitment, financial lending, and healthcare, where biased decisions can have profound consequences."
AIs are trained using data gathered from society and this information can often reflect societal and historical inequalities and stereotypes.
This means AIs could pick up these harmful biases without their developers even realising and they could become so deeply embedded they may be hard to detect.
It also warns that just removing attributes like race and gender from the data is an ineffective solution because the AI can often learn this information from other sources like names and where they live.
AI safety summit shows leaders recognise its dangers
The summit due to start next month at Bletchley Park, the site where Alan Turing's famous team broke the Nazi's Enigma code with an early computer, and is expected to attract leaders in AI from around the world.
Ahead of the summit, Sunak said he wanted to be "honest" with the public about the risks of AI.
He warned: "In the most unlikely but extreme cases, there is even the risk that humanity could lose control of AI completely through the kind of AI sometimes referred to as ‘super intelligence’."
In the speech, he also warned of the risks of AI falling into the hands of bad actors who use the technology to harm humanity.
The question of who should be the right people to use AI is already apparent with several Tory MPs, including former Prime Minister Liz Truss, criticising inviting China to the summit.
China should not be invited to the UK's AI Summit at Bletchley Park.
We should be working with our allies, not those seeking to subvert freedom and democracy.
I have written to the Prime Minister requesting that the invitation be rescinded.👇 pic.twitter.com/tQDe2P3kak
— Liz Truss (@trussliz) October 26, 2023
In a letter posted on X, Truss said: "No reasonable person expects China to abide by anything agreed at this kind of summit, given their cavalier attitude to international law."
Sunak defended the choice saying: "I know there are some who will say they should have been excluded but there can be no serious strategy for AI without at least trying to engage all of the world’s leading AI powers."
The Potentially Large Effects of Artificial Intelligence on Economic Growth, by Goldman Sachs