This interview is part of a new series of profiles on Charter called “Forefront” about leaders and thinkers on the forefront of the future of work. Subscribe to Charter’s newsletter and follow us on LinkedIn for the “Forefront” series about leaders and thinkers on the forefront of the future of work.
Ellyn Shook has a unique front-line perspective on the changes to work from generative artificial intelligence. As Accenture’s chief leadership and human resources officer, she is responsible for the technology and professional services company’s nearly 750,000 employees, many of whom perform tasks as part of their jobs that generative AI could take on eventually.
The scale is enormous. Accenture hires about 100,000 people each year, and last year said it spent $1.1 billion on learning and development for its workers over a 12-month period. It announced plans to invest $3 billion in its Data & AI practice over three years, including by doubling its AI workforce to 80,000 professionals. “You can think of us as not only a beta test for how you create a strategy around this, but how you execute it at scale,” Shook says.
At the same time, she confers with the company’s clients, with whom Shook says Accenture has done 700 projects around generative AI. “I’m super fortunate because not only do I get to learn through our own experiences, but I get to learn through our client experiences,” she says.
Shook’s belief is that “every single job is going to change” as a result of generative AI.
We spoke with her about the outlook for jobs—and about the perspective her own career provides—on the sidelines of the World Economic Forum’s meeting in Davos last month. Accenture had just released research highlighting worker concerns about AI’s impact on the quality and security of their jobs. (Charter Pro members can read more here about the most important research insights to emerge from Davos.) Here are excerpts from our conversation, edited for space and clarity:
You joined Accenture in 1988, and I’m guessing it’s one of your only post-university employers. What perspective does that give you?
It’s not my first, it’s my second. I worked for Marriott Hotels. I opened the New York Marriott Marquis. I was in a management development [program] and I was going through a catering sales rotation, and Accenture was my client.
It was 1988, so this is what they really said: ‘If you come and teach us service’—because I was in the service industry and I was serving them well, I guess—’we will teach you personnel. Come because our personnel people are not talking to partners, but you know how to talk to partners.’ That’s why I left Marriott and went to Accenture, which was Arthur Andersen at the time. I have been here since.
The biggest thing I’ll say—no one’s ever asked me this question before—is that when you’re someone like me whose parents were educators, learning is in every fiber of my being. And you’re given a chance to cross the threshold of a door or turn on a Teams call every day and learn. It’s the most enriching, vibrant career.
When you look over the course of my career, the opportunities that I had—like during the dot-com era, I worked for our high-tech incubator, and then when we were going to go public, they tapped me on the shoulder to help on the IPO team, and then I moved to France—every day, I was learning.
That’s advice I give people, whether they’re here, whether they’re elsewhere, or whether they’re here and want to go elsewhere: Just do everything you can to learn.
What is a personal experience that connects you to this moment of change for work and leadership?
Partly because of the research and learning through others, but really because of [Accenture CEO Julie Sweet’s] expectation that we deeply learn the technology, I was trying to understand how creatives are going to be impacted by generative AI.
I was talking to David Droga, [CEO of Accenture’s Song unit], and I said, ‘Are creatives really going to be out? Is Dall-E going to actually do the creative work or is the writing going to happen in ChatGPT?’ He said, ‘Let’s do an experiment, Ellyn. I know you love Hallmark movies.’ Which I do—I like happy endings. He said ‘Open Bard or ChatGPT.’
I typed in ‘Write a screenplay where there is a prince that lives in Europe and finds a girl on a trip in Australia’—which was the Danish prince story—’and make it romantic’ and all this stuff. In five seconds, it put a screenplay out. Then he took my computer and he added to the prompts, saying, ‘Make it in the style of ‘When Harry Met Sally’ and ‘An Affair to Remember’ and make sure that there’s a protagonist…’ Because he’s creative, he was able to really bring a lot more to the prompts than I would ever have thought to do as a non-creative.
He said, ‘That’s how creatives are going to be aided now.’ Going forward, this technology’s evolving. But he was confident about the power of generative AI to aid in the creative process and not replace the creative process. We have a lot of creatives here at Accenture and our Accenture Song business, and for me, that was a really important moment to see how their work is going to be better for them and how they’re going to be able to serve our clients better.
And the screenplay was good?
We got kind of into it and then we said, ‘Make it Oscar-award-winning.’ It got pretty good. It had a happy ending.
For someone like me, truly deeply understanding and not just saying, ‘Work is going to be elevated,’ but experiencing it is super important because you have to lead differently. You have to do the learning yourself. You would have seen in the research, almost two-thirds of people are worried about the stress and anxiety of losing their job. You have to have an enormous amount of empathy to be a great leader these days. You really have to understand how to bring people along in the journey.
What do you know from looking across Accenture teams internally about how AI is changing work and where the gains and biggest challenges are?
We’ve been applying generative AI into our core processes. For example, in sales, we applied it. We’re very data-driven here because we can be, and what I loved about that learning was that everything the research said would happen [about worker concerns] didn’t when they actually used generative AI.
Why is that? We didn’t just give them some sales chatbot and say, ‘Go do your job.’ They were the ones that reinvented the process. Involving people in the process brings down the anxiety level, it ups the trust level. When we actually applied what we believe—how do you leave people net better off?—to our own workforce, we saw the outcomes be very different than the fear of the unknown.
Is there any example, in terms of the usage of AI in your own workforce, where it significantly changed how teams work together and what their outputs are?
In the sales example, the job that they created from the process was very different and much more human-centric.
Like I said, we have massive amounts of data, and this goes back to the fundamental trust currency. Our people have full access to their data. They can opt out of certain things. Like if the skills inference engine says you’re a moderate skill level and you think you’re higher, you can take that out of the skills database.
What we said was, since you own all your data, we’re going to let the generative AI model—the model that my team and our technology team create—bring all of that data about what roles you played this year, what your priorities were, and how you delivered on them in a quantitative fashion. It’s called the People Leader Coach and it writes the first draft of the [performance] evaluation for [managers].
It used to take me 45 minutes to write one, and it wasn’t really that great. Now in 10 minutes you have your draft so that you can actually focus on the conversation, rather than like, ‘I’ve got to get these 15 reviews done.’
What does that mean to our people? You may have heard me say this for years: The more digital the world becomes, the more human connection is important. And this is allowing our people leaders, all 80,000 of them, to dial up human connection. That is a big change.
Your new research shows a lot of workers don’t trust their employer to ensure positive outcomes for everyone in the application of generative AI. Is it possible to change that, and how would you summarize what you need to do?
Yes. You need to involve your workers in the process. Instead of coming to them and telling them, ‘We’re going to give you this tool to do your job and it means 30% of you no longer have a job,’ you bring them in. Remember, there’s three components: How does the work change? How does the workforce change? And how do you prepare the worker? So you get them involved right up front in the work, because they know their job better than anyone else. They know it better than any leader. And getting people involved brings down that [lack of trust].
What do you think is the most underestimated challenge in shifting organizations to take advantage of AI?
I actually think it’s around human-centered change, because a lot of companies focus on getting the technology screwed in and that is not what’s going to change the company. In fact, I was talking to a CEO today who we’re working with on a large Salesforce implementation, and I know that they’re paying us to do the Salesforce implementation. I said to him, ‘Why do you think it’s not getting adopted?’ And it does not occur naturally to leaders that you need to invest as much, if not more, in the people around change than you do in the technology. That’s probably the biggest challenge I think that companies are facing.
Investing in the people around the change involves training. You have a massive training effort underway—one of the questions about training has been, can you do it at the scale and at the speed that’s necessary?
The training, I like to call it ‘learning’ because it comprises two things. Some of it is training, but the way people really learn is when you add it into the flow of work. So go train them: What is a large language model? How do you use it responsibly? What could go wrong? Learn that, and then introduce it into the flow of the work so that the deep learning happens. Deep learning doesn’t happen in a classroom. It happens from work-learn fusion.
You’re doing it at the speed and scale required?
It’s like anything: You have to have a strategy and an execution plan, and you have to be on your execution plan. For us, it sits in the transformation office in Julie [Sweet’s] office. It is a CEO-led initiative. In our enterprise navigator—which is not a dashboard because it’s AI driven, but you can think of it like a dashboard—she can see how many people are getting trained, is it happening. Large-scale change is a capability that we believe that all companies going through transformation need to build inside of their companies. We help clients do that.
Your research shows that 44% of working hours in the US are in scope for automation or augmentation using AI. How will that impact how we actually spend our work days?
One of the things you have to know is that there are new jobs also being created. One of the first things we did—because we are a technology company—is identify, we call it the R12, the 12 new roles that didn’t exist. A year ago December, they just didn’t exist here. Now we have 12 roles and we’re going to go from 40,000 people to 80,000 people doing those roles.
Julie Sweet has said that Accenture asks every job candidate, ‘What is something you’ve learned in the last six months?’
That’s how we test for something called learning agility. We believe that if people have learning agility and aspiration—we’re not brain surgeons at Accenture—that we can train you to do anything. Demonstrate that you learn. It could be playing tennis, playing bridge, whatever it is, and then that you have a desire to be successful, however you define that. So that’s why we ask people that. It’s super important. If people don’t demonstrate that they have learning agility, you can’t come work here because especially now, this technology came fast and it’s moving faster. The learning is going to have to be continuous.
What is something you’ve learned in the last six months?
I have learned to be a fairly basic prompt engineer.
Congrats.
Yes. How about you?
I’ve learned some advanced skills of parenting adult children and I’ve definitely learned some startup finance lessons.
That’s interesting. Sometimes the best way to learn also is to teach others.