Building expertise in the AI era

Hello and welcome to Working It.

I’m Bethan Staton, the FT’s deputy Work and Careers editor, standing in for Isabel this week.

Tomorrow is the longest day of the year. I was interested to learn this week that according to some readings, the summer solstice is linked to work. In early cultures, Midsummer celebrations may have marked changes between crop cycles, with rituals such as bonfires aiming to ensure the labours of the year paid off and brought a good harvest.

These days, shorter nights also change our experience of work: early sunrises might mean enjoying some extra time before heading to the office, maybe doing some exercise outside. Long evenings and warmer weather are an opportunity for after-work socialising: on Thursdays, at this time of year, the streets around the FT’s office are filled with the roar of pint-clutching, suited crowds that gather outside City pubs, the sound of after work in the summer time.

On the subject of ancient agricultural knowledge, read on for some thoughts from me about how more contemporary expertise might be in danger of being lost.

The end of expertise?

Deskilling — humans getting out of practice as technology replaces us at work — is a well-established risk of artificial intelligence. But in fevered predictions of how it will make our lives easier, or cause a job apocalypse, the more nuanced dangers of AI sometimes get overlooked.

That’s why I was pleased to review Matt Beane’s The Skill Codean anthropological take on how we develop skill, for the FT this month. Beane argues that to build expertise we need to start with easy tasks and gradually challenge ourselves by making them increasingly difficult, with the help of people we trust. Technology can disrupt that, by removing challenges or doing the early, easy stuff for us.

It’s pretty obvious how large language models might intensify that problem. I’ve already noticed many companies broadcasting the ways AI improves efficiency, often by doing work such as trawling legal cases, or basic research, that is typically done by novice hires.

I asked Beane what he thought LLMs might mean for skills and learning. He has bad news: the new technology looks like it will “make the learning bond more difficult”. By allowing us to skip over steps involved in a task, LLMs isolate us from “friction” activities — things such as looking something up, or asking colleagues for advice — that help us learn. Tech might help produce “slightly above average” work much faster, but this “reduces the challenge that would enhance your skills over time.”

That’s not to say these short-cuts are bad, nor that we should stick religiously with old ways of doing things. But it’s worth thinking about how people can continue to develop skills as work is automated. On that point, three pieces of advice Beane suggested stuck out for me.

First, if you want to keep learning make sure you are maintaining human connections: asking questions, working on tasks with colleagues, and helping less experienced workers. Make sure mentoring still happens. “If you’re not cultivating bonds of trust and mutual respect with other human beings, your skill development is going to hit a wall,” says Beane.

Second, push yourself, or make sure others in your team are able to. “The western world, to radically overgeneralise, has become a bit of a padded playground when it comes to work,” Beane says. Being “deprived of the ability to work close to your limits, to perform slightly less well than you normally would” means being deprived of ways to learn.

That segues nicely into a third point — keep in mind that the easiest paths come with pitfalls. Beane argues organisations tend to opt for productivity over skills development — making things easier or quicker rather than creating circumstances where staff can improve. “Maybe that gets you ahead in the short run. But what does that do to you in the long run?”

Of course, the ability to use AI well is in itself a skill. And some of the abilities we outsource to AI may not be greatly missed anyway. Expertise in crop cycles and bonfire building is of less use these days than when our ancestors were celebrating the summer solstice — even if some of us still get a lot of joy from the long days of midsummer.

Will AI stop us from acquiring the skills we need? I’d love to hear what you think — drop me an email at [email protected]

This week on the Working It podcast

What’s the state of AI in recruitment? Is talking to an avatar instead of a human interviewer going to help candidates overcome their nerves — and prevent hiring managers from falling for their own biases? I wanted to find out what the latest developments are in this (extremely) fast-moving arena, so in this week’s episode I talk to Ali Ansari, founder of micro1, home of the robot interviewer, and to Chano Fernandez, co-CEO of Eightfold, an AI-driven people management platform. I am still undecided on whether an AI recruiter is “better” — but it’s certainly faster. Isabel Berwick

Five top stories from the world of work

1. Interested in the WFH argument? Home in on the evidence: Do more flexible approaches to where we work improve performance, or make hitting targets more difficult? The results from this solid study on hybrid working are in — and if you enjoy a couple of days working from your living room, the news is good
2. ‘Insane’ pay rates for junior London lawyers raise concern over culture: I admit I’m not going to lose much sleep over junior lawyers paid £150k, but this piece from the frontline of the battle for talent among London’s law firms raises some important questions about what the trend for turbocharged remuneration means for culture and equality
3. London drags down UK productivity: London is still much more productive than the rest of the country, but recent Office for National Statistics figures showed the gap narrowing. That could be a sign of a less unbalanced economy, but it might also sound the alarm in terms of wider national growth
4. Making sense of Gen Z: employers seek answers on younger workers: How does Gen Z differ from other generations? It’s a question employers are asking more and more — and one a new crop of consultants and advisers are poised to answer. This fun piece delves into what they have to share
5. Employees embrace WFH (Work from hairdressers): Flexibility over where and when people work means a trip to the salon can now be extra productive, with some workers taking Zoom calls mid shampoo and cut. In response, enterprising businesses are setting themselves up as co-working facilities

One more thing

I’m currently reading Butter By Asako Yuzuki. Inspired by the real-life story of Japan’s “Konkatsu Killer”, it’s an entertaining exploration of the oppressive expectations placed on women, at work and in relationships, and a surprisingly perceptive take on life in journalism, too. Everything in the book is mediated through relationships with food, which means it’s introduced me to some delectably simple recipes — the book’s descriptions of rice with butter and soy sauce will have you salivating. Bethan

Next week is the first instalment of the FT’s webinar series marking the Business Book of the Year Awards. It’s on creating and energising a strong corporate culture — essential for anyone involved in talent retention or simply getting the best from your staff. You can sign up here.