The government must act to prepare for artificial intelligence (AI) to hit the workplace “like a freight train”, the boss of one of Britain’s leading energy companies has told Sky News.
Greg Jackson, founder of Octopus, says the adoption of AI across industry will ultimately improve the workplace and spawn new roles, but the startling pace of development means millions of jobs could be at risk in the short-term.
Octopus has seen huge benefits from the adoption of generative artificial intelligence in its customer service operations, with 44% of customer emails being answered, at least in part, by AI just seven weeks after it was rolled out.
Human employees still manage and check all the AI’s output, and Mr Jackson said it would not cost any jobs at Octopus.
He warned however, that the technology posed a threat to jobs at companies looking to cut costs, and business, regulators and politicians need to prepare for a rapid transition.
“Around the world, governments are quite quickly beginning to think about what they have to do but we haven’t got time to wait and see,” he said. “If a freight train is coming at you don’t wait to feel it hit before moving out the way.
“In growing companies, ones that are expanding and innovating in new areas, AI lets us do that faster, better for customers, and in our case hopefully better for the planet.
“But I think in companies that are not growing and don’t have the same opportunity to expand into new areas it could be a cost-cutting exercise in which case the threat to jobs is very real.”
“Right now we can see some of these impacts and I think responsible companies should be opening up this discussion so we can help governments think about how to handle it. And I think the first thing we need to think about is this economic dislocation and the risk to jobs.”
Mr Jackson’s warning comes as BT announced it will replace around 10,000 workers with advanced AI in the next seven years, making it the largest British company to make a direct link between the new technology and job losses.
The debate around AI has gained urgency in recent months with the emergence of new generative AI models such as ChatGPT and Midjourney, which can produce sophisticated written content and imagery based on a few text prompts.
The advances have surprised even developers, raising the prospect of a genuine industrial revolution in white-collar work, with the promise of productivity gains accompanied by fears of huge job losses.
While it’s not clear where the balance between promise and pain will eventually fall, companies are accelerating their use of the technology.
Allen & Overy, one the “magic circle” of major London-based law firms, began trialling a bespoke AI tool called Harvey last November which is now being used by 3,500 lawyers in 43 jurisdictions across the business.
Lawyers use it to generate a draft document or examine an area of law, which is then checked and finessed before being used, delivering productivity gains worth one or two hours a week, per person.
“It’s saving thousands of hours across a large organisation,” said David Wakeling, who has led the project for Allen & Overy.
“It’s a boring productivity gain, really, it’s an hour or two a week, but when you multiply that by three and a half thousand, that is a big deal for a business. It was impossible to find these productivity gains through a single deployment of a system.”
He said the technology was constantly surprising employees with its ability, but does not pose a threat to human workers.
“We see it as augmenting our lawyers, not replacing them… it is a brilliant productivity gain for some efficiency savings but the technology I’m seeing today, I’m aware that people talk about this [job losses] all the time, but we are using cutting edge technology and we are not seeing that impact today.
“We underestimate its capabilities all the time. Someone will send an email saying, I just got the most amazing answer or I just found this use-case, it still happens a lot.
“It’s still limited, it still has the risk of errors, we still have to concentrate on making sure it’s safely deployed and people understand that you need the expert in the loop. But fundamentally, it’s an amazing machine and it produces surprises all the time.”
Concern for workers’ rights
While employers search for opportunities in AI, unions are concerned at its potential to erode workers’ rights and are calling for tighter regulation.
The government wants the UK to be a world leader in AI, and in a recent white paper said it would not legislate to deal with AI, preferring to allow existing regulators to work with companies on appropriate rules.
The TUC says workers are already under-represented in the rollout of new technology and is calling for legislation to protect humans from hiring and firing by algorithm.
“Our research has found that unfortunately, there’s a very low level of consultation at work about the introduction of new technologies, and indeed, sometimes technologies are operating and making decisions about people who don’t even know that that’s happening,” said Mary Towers, the TUC’s lead on AI.
“We say that at the very moment at which regulation is most needed, when the technologies are developing so rapidly and the implications are so significant, instead of regulating, the government is putting forward flimsy and vague proposals that don’t have any statutory footing.
“There’s potential for everyone to benefit from the innovation and from the development of AI-powered technology, but the critical issue is, are lots of different voices represented at the development stage of the technology?”