AI and work

With every leap in AI’s capabilities, a question nags at the back of many people’s minds: how long before it takes my job? Increasingly, AI is becoming ever better at writing, editing, coding, diagnosing, completing tax forms, parsing legal documents and doing most things that have always been the preserve of white collar workers. And, crucially for employers, it’s cheaper to run and works 24/7, never asking for holiday time and sick leave or struggling to produce PowerPoints with a hangover.

Little wonder then that some leading lights of the AI tech fraternity are telling desk jockeys to retrain as plumbers.

But white collar workers shouldn’t be too gloomy about their prospects, argued Dr Mostfafa Ajallooeian and Andreas Blum, from Swiss data services company Unit8, in a recent presentation to Pictet Group.

“If people acquire skills, their job functions change,” says Ajallooeian, Unit8’s engagement director – and AI will encourage the development of these new skills. He adds that adoption of AI means “efficiency is going higher, but you have new things to do.”

Ajallooeian points to Goldman Sachs research which notes 60% of jobs that currently exist weren’t around 40 years ago, with the development of modern technology driving the creation of those jobs. After all, there’s no longer large numbers of blacksmiths and chimney sweeps, but that doesn’t mean people who would otherwise be employed making horse shoes or sweeping chimneys are jobless.

The times they are a-changin

That there will be changes is undoubted, says Blum, head of data and AI consulting at Unit8.

“Back in the 19th century it was the steam engine that revolutionised how our society works. Now we’re at another stage… but this time it’s the GPUs (Graphics Processing Units) that are powering this new revolution,” he says.

One corner of industry where this revolution is starting to take effect is financial services, according to Blum. AI is already being used to automate compliance checks, summarise reports, and provide client insights.

“It’s clear (AI) is going to have a quite large impact (with) productivity gains of between 20% and 30%,” he said. “Jobs are not disappearing at a grand scale, but jobs are changing and maybe with the same workforce.”

Employees will be able to serve more clients, resolve issues faster, and focus on higher-value tasks, as AI handles routine work. AI is positioned as a tool to augment human work, not simply replace it. For example, the technology can draft documents or analyse data, but humans are needed to provide oversight, critical thinking, and domain expertise.

AI also enables businesses to undertake activities that were previously too resource-intensive or unprofitable, opening up new revenue streams and business models.

But at the same time the adoption of AI increases the need for workers to gain new skills. They will need to develop basic data literacy and analytical skills to work effectively with AI tools. There is likely to be a growing demand for “citizen data scientists” and the use of low-code/no-code platforms to democratise access to data and AI.

There are downsides as well, however. Over-reliance on AI tools could lead to a decline in certain skills as people offload dull tasks to machines. Already, there is concern among legal professionals that replacing much of the drudge work that junior lawyer do with AI will not only shrink the pool of talent at the bottom reaches of the profession, but also strip it of the skills that are essential to lawyers at the top of the industry – those who can’t be replaced with AI.

And, finally, building AI into systems raises governance and security challenges, particularly around data access, privacy, error management and the correct use of AI outputs. Companies need to structure frameworks to manage these risks.