How to do AI well

Hello and welcome to Working It, the FT’s weekly guide to the workplace and leadership.

This week, I’ve been reading my decades-old diaries 📚 (an FT column is coming) — but also thinking about the future, via two interesting new books. In Alastair Campbell’s Persevilience, published next week, the podcaster and political adviser coins a word that, he believes, reflects a better way to thrive in life and work. (The word is a blend of perseverance and resilience with a dash of curiosity and optimism.)

The second book, Herminia Ibarra’s The Bonus Years: Your Next Chapter in Life and Work isn’t out until January, but I’d strongly recommend pre-ordering if you are thinking about “what next?” 🧐 Or if you are already going through career/life transitions. I loved it.

PS We interviewed Alastair at the FT for the next Working It film — coming at the end of this month.

Read on for CEO insights into the latest developments in hybrid human/AI workforces 🔮; and in AI Clinic, a futurist offers tips on getting more “friction” into your workday.

What makes for a successful AI transformation?

AI spending is rocketing 🚀 — the average UK business is spending £20.6mn this year, up from £16.2mn last year, according to new figures from an SAP/Oxford Economics report. And yet, a lot of the rollout (40 per cent in this survey) is “piecemeal” or department-led, rather than part of a company-wide strategy.

At an AI conference last week, I heard Rich Klee from Palladium Digital talk about “AI vertigo” — a term new to me, but it seems to have been coined in 2023. It describes a pace of change that’s so dizzying that it makes people feel overwhelmed 😨.

This sums up how many workers currently feel about AI, so I was keen to hear from CEOs who can give examples of how they have overseen successful, system-wide change, both internally and in their customer base. Denis Machuel is chief executive at Adecco Group, a big recruitment company and technology consultancy, and Brian Holliday is UK and Ireland CEO at Siemens, the industrial and manufacturing infrastructure company.

I talked to Denis and Brian at an FT event this week, joined on stage by Anna Thomas, co-director at the Institute for the Future of Work, a research organisation. We covered the current state of global AI rollout and barriers to scaling beyond superficial change, as well as their experiences of teams where a hybrid human and agentic workforce is already embedded 🦾.

Anna suggested that leaders first go back to basics, by looking at how roles and processes are designed in organisations. “It’s a really good way of understanding what’s known, breaking it all down.” That basic mapping of roles and requirements isn’t happening enough. If now case studies suggest it’s a “wild west” in terms of whether companies map their current work design before they start to implement change around AI.

Brian echoed that point, saying that Siemens’ advice for a long time has been “don’t digitise it, understand how to improve your process first, otherwise all you have is a bad process. So exactly the same logic applies to AI. Where is your firm trying to create value? Where are you going to go for the KPIs you are trying to solve? And only then start thinking about how we’re going to apply technology.”

Once that’s done, Brian said, you can start thinking about “the art of the possible”, and he focused on physical AI, used in many factory settings. “How do you deploy globally, how do you take people with you? These are leadership topics.”

Denis said trust and good leadership are vital for AI transformation to be effective. A recent Adecco Group survey found that those messages are often not getting through: 36 per cent of workers, mainly in white-collar sectors, “told us that their company had explained to them what was really happening and how they were envisaging the impact . . . and 39 per cent said they were involved in the way their work was being impacted and reinvented.

“At some point all of us will be faced with agents, working with robots, and all of that needs to be properly orchestrated” in order to maximise what humans and AI can do together. He described the rollout of agents in his recruitment business as a demonstration of a hybrid human and agentic process. Leaders from the recruitment business, IT, HR and finance came together to design the system, and “we broke the jobs down into tasks, into micro-tasks; it’s by discussing with recruiters that you discover how things happen in real life.”

Once the processes were understood, the teams decided where to use AI and what should remain with humans. It’s key that everyone from different teams works together: “Be super concrete on the work, on the way it happens, so that people realise it’s multi-dimensional and that only together [will] they solve it.”

The importance of ongoing, open conversation was shown when Denis visited a group of French recruiters and discovered they were using WhatsApp to talk to candidates because it was easier to use than Teams when they were travelling. “If you don’t understand that, through constant conversations, you miss it — and think, because you have Teams and the proper tools, then you’re done with AI.”

• In a nutshell: Do the basics first in terms of mapping internal design and processes — before you start rolling out AI across businesses. You may get widespread adoption, but without real absorption and change.

In a nutshell: Do the basics first in terms of mapping internal design and processes — before you start rolling out AI across businesses. You may get widespread adoption, but without real absorption and change.

• Want more? Ethan Mollick, Wharton School associate professor and AI expert, demonstrates how quickly AI systems “learn to seek out information themselves”. His Substack, One Useful Thing, is a must-read for leaders.

Want more? Ethan Mollick, Wharton School associate professor and AI expert, demonstrates how quickly AI systems “”. His Substack, One Useful Thing, is a must-read for leaders.

AI Clinic 🩺

A break from problems this week (there’s enough gloom around ☁). I asked the futurist Diana Wu David, director of ServiceNow Futures and founder at the Future Proof Lab, to tell us how she uses AI — and offer some tips:

“At work, I use it to introduce friction. As a futurist, my job is to see around corners and question assumptions, not just find answers. I often ask AI to argue for other scenarios and against my own, or surface what might make my conclusion or point of view wrong.

This ‘AI as challenger, not assistant’ approach is useful because most of the organisations I speak to aren’t struggling to generate more ideas. They are struggling to improve decision quality and speed of insight.

One fun way to do it is to make an [AI-generated] personal board of advisers. For example, the CFO, a sceptical consumer, a transformation guru, a historian and a futurist. Sometimes I include specific people or mentors. I ask them to critique and debate from their point of view. This surfaces perceptions I don’t have easy access to and lets me throw out 90 per cent of my ideas — and keep and refine the rest.”

Got an AI problem, dilemma — or a tip on innovative uses at work? Email: isabel.berwick@ft.com. Problems are anonymised.

Five top stories from the world of work

1. Mark Zuckerberg says Meta has reached AI ‘frontier’: Hannah Murphy has an in-depth interview with the tech billionaire; I was struck by his assertion that Meta’s AI work is “focused on personal empowerment” and wanting to “build things that are good in the world”.
2. HR doesn’t need to be the CIA: Sarah O’Connor on the new cyber security checks coming to the recruitment process in a bid to foil fake applicant profiles and automated applications.
3. How AI tempts the lazy mind: My favourite read of the week. Tim Harford’s column examines how easy it is to let AI take the cognitive load, and gives a neat example of what happens when we let it do just that.
4. City of London hits out at Andy Burnham’s plan to scrap non-compete clauses: The proposals are welcome at tech firms but not at some hedge funds and banks, which rely on non-competes to stop star performers being poached. By Martin Arnold and Kieran Smith.
5. Why Infantino is so hard to topple: A great analysis from Simon Kuper of why the head of Fifa is still in his job, despite multiple scandals. He is even likely to win re-election to head football’s governing body.

Mark Zuckerberg says Meta has reached AI ‘frontier’: Hannah Murphy has an in-depth interview with the tech billionaire; I was struck by his assertion that Meta’s AI work is “focused on personal empowerment” and wanting to “build things that are good in the world”.

HR doesn’t need to be the CIA: Sarah O’Connor on the new cyber security checks coming to the recruitment process in a bid to foil fake applicant profiles and automated applications.

How AI tempts the lazy mind: My favourite read of the week. Tim Harford’s column examines how easy it is to let AI take the cognitive load, and gives a neat example of what happens when we let it do just that.

City of London hits out at Andy Burnham’s plan to scrap non-compete clauses: The proposals are welcome at tech firms but not at some hedge funds and banks, which rely on non-competes to stop star performers being poached. By Martin Arnold and Kieran Smith.

Why Infantino is so hard to topple: A great analysis from Simon Kuper of why the head of Fifa is still in his job, despite multiple scandals. He is even likely to win re-election to head football’s governing body.

One more thing . . .

In Sharon Horgan’s excellent comedy, Youth, on HBO Max, Horgan plays 50-something Alex, bringing up teenage Ruari on her own, with occasional appearances from his useless father. Alex works as a literary agent (an author friend tells me the publishing industry details are very accurate) and Alex’s zero-boundaries relationship with boss/best friend Sam (Sharlene Whyte) is stifling and complex. The FT gave it five stars.

A word from the Working It community ✍🏻

The recent newsletter about CEOs turning to philosophy and academic courses 🏛️ for their leadership training continues to pull in interesting responses, including from Helen Beurier. Inspired by “salons”, where people gather to discuss big ideas, she is putting on a “Humans at Work” festival in London tomorrow (Thursday). I asked her to tell us more:

“After 30 years in HR, I am transitioning from chief people officer to philosopher as I pursue a doctorate of philosophy. Why? While technology continues to extend the limits of our achievements, we risk forgetting how we best unleash the maximum potential of humans themselves. “Ancient wisdom, alternative thinking, offers us a path back to the future. Humans are better together. I am bringing the concept of the ‘salon’ to life to revive the art of conversation for human flourishing. Salons are designated spaces where your contribution to conversation matters more than your roots, exploring topical issues to strengthen our ability to think critically and creatively for wiser choices and greater resilience to change.”

“After 30 years in HR, I am transitioning from chief people officer to philosopher as I pursue a doctorate of philosophy. Why? While technology continues to extend the limits of our achievements, we risk forgetting how we best unleash the maximum potential of humans themselves.

“Ancient wisdom, alternative thinking, offers us a path back to the future. Humans are better together. I am bringing the concept of the ‘salon’ to life to revive the art of conversation for human flourishing. Salons are designated spaces where your contribution to conversation matters more than your roots, exploring topical issues to strengthen our ability to think critically and creatively for wiser choices and greater resilience to change.”

And finally: join us at the FT to talk about better workplace culture 😍

Is your culture built on innovation, or internal competition? In most organisations, it’s the latter. Ruchika T Malhotra’s new book, Uncompete: Rejecting Competition to Unlock Success, is a blueprint for change. I’ll be in conversation with Ruchika at FT HQ in central London on the evening of November 3. We’ll discuss how removing internal competition can improve performance, inclusion and psychological safety. Ruchika will also talk about her experiences guiding leaders and staff to move beyond the “winner takes all” mentality.

It’s a free event, and we’d love Working It readers to come 👋. Register here.