ADI IGNATIUS: Welcome to the HBR IdeaCast. I’m Adi Ignatius.
Artificial intelligence promises to make us faster and more productive, but there is an underlying question as to what it means to allow AI to do more of our thinking. What will happen to our ability to learn, to improve, and develop expertise over time? What will happen to our ability to learn, to improve, and to develop expertise over time?
Today’s guest argues that the answer depends less on the technology itself and more on how we choose to use it.
Angela Duckworth is a professor at the University of Pennsylvania and her research has long focused on what separates high achievers from everyone else. Enter: AI. I spoke to her recently as part of the 2026 HBR Leadership Summit about her research that suggests AI can actually accelerate learning when it’s used as a coach rather than a crutch.
We’ll also hear about why young workers feel both excited and uneasy about these tools, and we’ll hear practical advice for leaders who want their teams to become more productive and more capable.
Duckworth is the author of Grit, as well as of the forthcoming book, Situated: Find the People and Places That Bring Out Your Best.
ANGELA DUCKWORTH: As you know, I do study grit, but I want to put grit in the context of AI and I want to cover three questions. One is how grit is integral to becoming excellent at what you do, no matter what it is that you want to do, and how that relates to the second question, which is AI and whether AI helps or hinders grit, and under what circumstances a leader might make AI work for you versus against you. That, of course, is the third question in particular, which is how a leader and a team might optimally use AI to become better at what they do.
So to begin, what is the relationship between grit and excellence? Why is grit a necessary, if not sufficient, element to successful leaders, successful individuals in any domain that I have studied? I’m going to give you a little test. So you can take it right now. It’s got four questions on it. And you’re not going to formally score yourself, but these four items are from the grit scale. It’s a questionnaire I developed from interviewing Olympic gold medalists and Nobel Prize winners, world-class achievers across domains.
And here’s one item. “I am a hard worker.” If I ask the people who know you best, how many would go out of their way to say that you are an especially hard worker? The more that is true, the higher your grit score. And in particular, the higher your perseverance sub-score, because grit is not just perseverance, but it’s also passion. So let’s start with, “I am a hard worker.”
Here’s another perseverance item on the grit scale. “I finish whatever I begin.” High achievers who have grit have a bias toward completion. They hate to quit. They especially hate to quit when things are hard and everybody else is quitting. So these two items index your perseverance for long-term goals. That’s half of grit.
Let me show you two other items you can reflect on your own character at the moment. Now, before I show them to you, I need to warn you they are reverse-scored. And by that I mean, when I show you these items, if you say, “Wow, Angela knows me. These are just like me,” you actually get a lower passion sub-score and a lower overall grit score. “I have difficulty maintaining my focus on projects that take more than a few months to complete.” Here’s another reverse-scored passion item. “I often set a goal, but later choose to pursue a different one.”
When you ask me about the passion of high achievers, I want to point out that it’s not about intensity. It’s not about pulling all-nighters. It’s not about being an extrovert. It’s not about having manic levels of energy and enthusiasm. Excitement is common, but endurance is rare. The passion of gritty high achievers is consistency. It’s being loyal to some top-level goal that keeps you interested, keeps you committed over long periods of time. And that is why grit is not just perseverance, but passion and perseverance for especially long-term goals. Weeks, months, years, decades, in some cases, a lifetime.
Now, this diagram is an X-ray of what goes on in the mind of a very gritty high achiever. They do have a top-level goal. They have an ultimate why that gives meaning and sense and direction to all of their mid-level goals, things I want to accomplish in the next five years or three years. And then you go down, this is, of course, a simplified diagram, but at the very bottom of the hierarchy is your to-do list. The very bottom is what you have to get done today and this week.
And here’s where I think you need to be agile, adaptive, and respond to circumstances. So you should be stubborn at the top. Your value shouldn’t change, your mission shouldn’t change, your top-level goal should not change, but very agile at the bottom. And you have to be agile in part because the world is changing. So you want to capitalize on those opportunities, including AI. And all of this keeps you on a learning curve.
This is a slide that, stylized fashion, summarizes the research of Anders Ericsson, a cognitive scientist who coined the term the 10,000-hour rule. 10,000 hours of practice. What Anders studied was the learning curves. We have skill on the vertical axis, you have time on the horizontal axis. He studied the learning curves of people who became grand chess masters or prima ballerinas or world-class CEOs. And what he found was that if you can stay on a learning curve and continue to improve, then maybe even more important than talent, he would say that, I would say that, is your ability to consistently work hard, not give up when everyone else is giving up, and actually stay focused on something long enough to get really great at it.
And this is why it’s so relevant to AI, because the question that looms I think in everyone’s mind is, will individuals continue to become more excellent themselves or, with an invention that does your thinking for you, will just the opposite happen?
I want to share with you data that was collected in collaboration with the Walton Family Foundation and Gallup. This is a nationally representative sample of Gen Z young adults, 18 to 28, about 2,000 of them. And we asked them in late October of last year, “How many of you have used AI in the last month for the following purposes?” And you can see that three out of four young adults have used AI in the last month as of October. That’s probably north of that figure today, which is just a few months later. The number one use is Google or a substitute for Google.
Essentially, if you look at this rank ordering, what you see is that Gen Z is using AI primarily to be more productive. You might be alarmed at those numbers at the bottom of the figure, using AI as your girlfriend or boyfriend or as a substitute for a friend or when you’re not supposed to. We can certainly discuss that in the Q&A, but primarily I want to say that Gen Z is leaning in to using ChatGPT, Claude, and other platforms to become more productive.
In the same survey, we asked them how they felt about AI use. And there’s a deep ambivalence. They love that it makes them more productive, but they worry, and they worry about the same things that you worry about and that I worry about, which is that if I use AI, it’s like a crutch and it’s going to make me as an individual over the long-term lazier and less smart.
And that brings me to the question of, what to do? As a leader, what to do? If we understand that grit keeps you on a learning curve, if we understand that excellence takes a lot of hard work and continuous improvement, but now we have this invention which makes us definitely more productive in the short-term, potentially at long-term cost, what hope is there? And I want to share some research findings that do give me hope.
I ran an experiment. This was led by my then graduate student, Benjamin Lira. We randomly assigned individuals to condition, but everybody in the experiment had to rewrite a terrible cover letter. So if you look on this laptop screen, there is this long, oh my God, I don’t want to read it, cover letter to a hiring manager. We gave that to the young adults and others in our study. And we said, “We want you to rewrite it.” But in one condition, you had to do it on your own. In other conditions, you were allowed to use AI to do your work for you. So that would be the simulation of what a lot of people are doing today, which is copying, pasting something into an AI platform like ChatGPT and having, at the press of a button, an instantly rewritten cover letter. And then we had a holdout control that did no practice at all.
And I want to share the results. Imagine that you were in this experiment. You either had no practice at all or you had to practice on your own without an AI tool, or finally, you were in the condition where you got the tool. And if you predict the results of this and you are like the people that we sampled in a separate forecasting study, you would predict that over time, the people who use the AI tool would actually maybe write better cover letters in the moment, but on a test, their own writing skill would be diminished. This is the concern that AI makes us lazy and stupid.
And what we found, in fact, to our surprise honestly, was that when you see this cover letter that your AI chatbot has produced for you instantly, you actually yourself get better at writing skill, not just immediately, but a day and a week later. So your writing skill has enduringly improved. You are better than practicing on your own. You are also better than the participants who had no practice at all. In other words, improbably, we found that using AI to boost immediate productivity in this case also boosted long-term skill. Not at all what we expected and I think not at all what other people expect.
And here is what I think is going on. When you use AI, you may reduce your effort in terms of doing things faster, not thinking as much as you would if you had to do it all on your own and struggle. So that’s effort. And so we can predict that using AI use would reduce effort. In our experiments, that’s exactly what we found. During the practice period where you had this AI tool, you used fewer keystrokes, you spent less time. So yeah, when you’ve got an assistant, you will work less, and that should diminish your skill. I think everyone is correct to worry that AI use is going to make them lazy and stupid because all things being equal, if you reduce your effort, you will learn less. You will improve less.
However, I think what we’re forgetting about is a second box on this figure, and that is the quality of your learning environment. Let me give you an example. If you are Roger Federer and you are a young boy with lots of talent, imagine that you had a world-class coach, which he did. He did have a world-class coach who helped him realize his potential. Now, imagine you’re the same Roger Federer and you have a terrible tennis coach, somebody who can barely play tennis themselves, knows nothing about motivation. Now, that is going to degrade your learning environment.
My point is that it’s a open question whether learning environment will be improved or diminished by AI use. But I think in our research, what we find is this. When AI can do something better than you can do, when AI can format a cover letter better than you can do, it increases the quality of the learning environment. And how does it do that? In the same way that a great coach does, which is to show you by example what you need to do. When Claude writes better than you, when ChatGPT comes up with an idea that you didn’t come up with, watch and learn. So much of human learning is by example. And therefore, I want to say that there is hope because even when we reduce effort, if we improve the learning environment enough, we can see improved skill over the long-term, which is what we found in our experiment.
The great coach, John Wooden, who led UCLA to a record 10 NCAA basketball championships, boiled down learning to this little poem. Number one, explanation. Number two, demonstration. That’s what I’m talking about in terms of learning by example. Number three, imitation. So his players would see him do a move and they would copy it. And then repetition, repetition, repetition, repetition, repetition. I argue that in the age of AI, it is possible for your teams to get smarter and maybe even grittier if GPT or Claude is explaining what to do, demonstrating what to do, if we lean in and say, “We can watch and learn,” and then, of course, if we stay on the learning curve enough to become excellent at what we do.
This is why I think it all comes back to having a steadfast goal, a top-level goal that you’re sticky about, but being agile and adaptive at the bottom, including embracing ways of using AI. In some, grit matters because it keeps you on a learning curve. And you will never be great unless you keep climbing some learning curve in life. Number two, young workers today are using AI to boost their productivity, but they’re ambivalent. They worry that it makes them lazy and stupid. And finally, when AI does something better than you, learn by example. And I argue, be willing to outsource to AI the goals at the very bottom of your goal hierarchy where, frankly, you’re not learning at all, you’re just completing some chore or task.
ADI IGNATIUS: Lots to talk about, lots of questions coming in. I do want to ask one quick question of my own, which is, and I should remember this because I’ve heard you say this before, but you talked about grit, long-term perseverance, a need to finish what you start. How many of us have that?
ANGELA DUCKWORTH: Well, Adi, I will say this. Like most of human-
ADI IGNATIUS: Is that a rare quality or is that a common attribute?
ANGELA DUCKWORTH: … qualities, it’s not either/or. It’s how much. So you can be very, very, very gritty. You can be just very gritty. You can be sort of gritty. You can be not very gritty. You can be extremely ungritty. So it is a continuum.
I do think that when you look at the Michael Phelpses of the world, the Katie Ledeckys, or the titans of industry and finance, I mean, almost always they are outliers in grit among other things, by the way, because I don’t think grit is enough. It is necessary, it is not sufficient. And so wherever you are, you’re on the continuum, Adi. I am too. And what I will say is this, if there’s one singular insight that should give you hope, it is that wherever you are in the continuum, we find that grit changes. And in particular with age and with experience, on average, people get grittier.
My argument is that if you understand what is in the head of a gritty person, if you understand goal hierarchies, for example, and the importance of a top-level goal and the importance of being flexible at the bottom of a goal hierarchy, you can catalyze. You can accelerate the development of your own grit.
ADI IGNATIUS: You talked about the question of whether AI is going to dumb us down or weaken our skills and ability. And you talked about John Wooden, the coach example. The coach shows you what to do. You learn from that. But AI, it almost seems like the coach doesn’t just show you once how to shoot a layup. It does it for you for eternity.
So here’s one of the questions that came in. This is from Begum Avdagic, product manager at Dewhurst UK. If grid is about sustained effort toward long-term goals, is offloading this cognitive work to AI undermining that muscle you’re trying to develop?
ANGELA DUCKWORTH: I want to agree with this concern, which is that if you continuously offload all cognitive work to AI… I mean, when I say that the human brain and the human body are adaptive, I really mean it. And it goes in both directions. I mean, anybody who’s had their arm in a cast is just shocked to see how quickly it atrophies. So we are plastic. We can grow. We can grow smarter. We can grow dumber. We can grow more gritty. We can grow less gritty. So it’s not a guarantee that the gains are always going to be monotonic. So I think it’s a terrific concern.
But let me give you the example of the graduate student who led that research. When AI debuted and changed everyone’s lives, he was one of the very first to realize that. He walked into the door of my office. He said, “I’m dropping everything and I’m going to study AI.” I didn’t even know what he was talking about. Then actually in the next few months, he got a little discouraged. He was like, “I spent my whole 20s learning how to program, and now I cannot program as well as GPT.” And those were early models by the way. So there was a stages of grief. It was like, “Oh, I just trained my whole life to do something and I cannot do it as well as this mindless bot.”
Then what happened was two things. One is Ben learned to delegate the kind of tasks that really there wasn’t much of a learning curve. He just had to get them done. Really tedious tests, formatting tables, putting the numbers into a manuscript that we were going to publish and then editing it and making sure that it was in APA format. He delegates those tedious tests to AI. So he’s not getting better at them, but it doesn’t matter. He has freed up hours and hours of his day to do the kind of high-quality work at the top of the goalie hierarchy that he really wants to do.
And as I said, he does learn from example. When he asks Claude or GPT to do some coding for him, which is what he does, because we’re, of course, all programming that way, he can see the code. And sometimes he’s like, “Oh, God, why didn’t I think of that?” So I think a lot of it boils down to how you choose to use AI. If you are going to cognitively offload everything and choose not to learn and improve, then you will become lazier and stupider. That’s absolutely true. But if you choose to free up time so you can do higher-level thinking and every time you use AI, if you ask yourself, “Could I learn something from this? Could I myself, as a human, get better?”, then I think the opposite will be true.
ADI IGNATIUS: Laziness and stupidity is a choice.
Okay, so here’s a question from Satoi, who is a graduate student at Harvard who says, “Hi, Angela. I’m a huge fan. Thank you for presenting. When you say watch and learn, how do you ensure that your team isn’t just watching and moving on, but actually taking the time for it?”
ANGELA DUCKWORTH: I do think you need to have a mindset about this that some would argue is a maximizing mindset. So the great Nobel Laureate, Herb Simon, who is no longer alive, but he taught for years at Carnegie Mellon, and he is really a giant. I mean, a giant even among the Nobel Laureates. And he distinguished between maximizing and satisficing. And in a lot of life, you want to satisfice, which means good enough is good enough. You go out to lunch and the chicken salad sandwich is eh, but you’re not going to ruminate about it all day. It’s like, “It was good enough. We had lunch. What are we doing this afternoon?” And you can make the argument, David Epstein makes this argument in his latest book, Inside the Box, that that’s exactly what you should do for many things.
But as somebody who has studied Olympians and somebody who’s studied three-star Michelin chefs, and as somebody who has studied really the world-changing CEOs, I will tell you that if you commit to excellence, there is a domain in your life where you say, “I am not a satisficer. I am a maximizer. Good is never good enough.” If you have this mindset, if you say, “As a psychologist, I want to become a better psychologist,” if as a manager, you want to become a better manager, as a strategist, I want to become a better strategist, then every time you’re interacting with AI, you’re asking yourself, “How is this going to improve me?” I’m delegating tedious tasks. I don’t have to do that anymore. I’ve got four hours to do something really hard and important. Great. Or I watched and I learned. So it’s not just what you’re given, it’s what you’re taking from this interaction with this new technology.
ADI IGNATIUS: Yep. All right, here’s a question from someone identifying themselves as Dr. Indira, CEO and founder at EmpowerU. And the question is, it’s interesting, from an AI ethics and human-centric governance perspective, when younger workers express ambivalence about AI, it might be less about whether or not they have grit or whether they’re open to change, but more a rational response to the accountability gaps when you don’t know who is responsible when AI is wrong. So how should leaders address that kind of implicit structural dimension besides the motivational one?
AI ambivalence may not just be, “Come on, you’re not changing with the times. You don’t have grit. Let’s get moving here,” but maybe a more rational wait-and-see approach to a technology that isn’t proven, that makes mistakes, has accountability gaps. So it’s saying you’re not getting with AI… It may not be just about grit –
ANGELA DUCKWORTH: And by the way, Adi, I don’t think I ever said, “Get with the program. Be gritty.” By the way, that helped nobody ever. I have two daughters. I never once said, “Amanda, be gritty.” I never once said, “Lucy, be gritty.” I actually think that’s like… Bumper sticker psychology doesn’t work. I don’t think that’s the message of this talk. So the message of the talk is grit is about being devoted to a top-level goal and being flexible about how you get there.
But I will say to this excellent question, if I interpret it correctly, I think most of human responses are rational. Why do I think that? It’s because people are not waking up in the morning and thinking, “I’m going to make a logical mistake.” Now, do people make mistakes? Absolutely. But almost always there is something that you’re responding to in a sensible way, even if it might be incomplete or myopic.
So when people have this hesitation about AI, I think it is for good reason. I’ve hoped to have made that point that yes, absolutely AI can make you lazy and stupid. I also think it can make mistakes. So yes, hallucinations are still possible. And more than that, when you have this giant genie that does auto complete, it’s like playing hangman. It’s like C-A-T, C-A-N, more likely than C-A-X. So this thing that we don’t fully understand ourselves that’s magical in its abilities, but we also have a very different way of thinking. Human beings don’t think with auto complete. We should be hesitant. That’s a rational concern, not just because of lazy and stupid concerns, but also because we don’t fully understand how this genie works.
So I think all those things are true, and yet I am among those who say, “If you don’t wake up every day and ask yourself, ‘How can my company use AI to improve,’ then you are an ostrich with your head in the ground.” The hesitance is itself a decision. The hesitation is itself an action. And so we don’t really have a choice. When Ben Lira walked through my door and said, “I’m dropping everything I’m doing and I’m working on AI,” I couldn’t believe it. And I though he might be wrong, but now in 2026, I think he was exactly right.
ADI IGNATIUS: This is a question from Simon, who works on HR tech strategy and transformation at BHP. And he asks, he says, “Your concept of situational agency, designing environments rather than relying on willpower,” the question says makes intuitive sense at an individual level, but what does it look like when leaders try to apply it at scale across a big organization and where can that framework break down?
ANGELA DUCKWORTH: I spent the first half of my life answering a question which is why are some people more successful? And I looked the first place you would look, which is you look at the person and you’re like, “What are their personal qualities?” And you notice that Michael Phelps is really gritty. You’re like, “Whoa, this kid didn’t miss a practice for 10 years.” And by the way, when Michael Phelps was training, he didn’t train five days a week. He didn’t train six days a week. He trained seven days a week. So you’re like, the answer to the question, what makes some people more successful than others, is the personal quality of grit. That was the first half of my career.
The second half of my career, I’ve been trying to answer the same exact question. What makes some people more successful? But instead of looking inside the person for their personal qualities, I’ve been looking without. I’ve been looking outside them. Who coached Michael Phelps? Where did Michael Phelps train? What kind of team was he on? What culture was he in? And that’s what I mean by situational agency, your ability to change your environment in ways that bring out your best.
Michael Phelps had a coach from when he was 11 to when he retired at 31. That coach is Bob Bowman, the best swim coach in history. If Bob Bowman were a country, he would rank pretty high in medals for swimming at the Olympics. He had the best coach in the world. Where did he grow up? He grew up in Baltimore, 10 minutes away from the premier aquatic youth center in the world. 10 minutes away. Two 50-meter pools, one inside, one outside, and world-class competition. How many of you can say, “Oh yeah, I also grew up there”?
What kind of team was Michael Phelps on? Michael Phelps was on a team where people went to the Olympics all the time. They broke records all the time. And finally, what kind of culture was Michael in? Well, I’ll tell you, he had a culture at that team that Bob Bowman created, a culture of excellence, as Bob Bowman said, and he also had the world’s best mom. Debbie Phelps raised her three children in a culture of excellence.
So when I talk about situational agency, I say, “Well, some of that’s luck, of course, what your situation is, but some of it you create.” Michael followed Bob Bowman when Bob changed locations. He followed him because he knew he had the best coach. He always navigated to be around the best teammates. He tried to not spend time with people who didn’t bring out his best. He always cared about the quality of his physical training facilities.
So my last word to you, Adi, is that grit is important, but so is your situation. And if you want to be your best, you not only have to have passion and perseverance for a top-level goal, you also have to be deliberate about your physical situation, where you keep your phone, about your social situation, who you are working with on a team, who your mentors are. And of course, ultimately what every leader worries about, you have to be very intentional about culture. I don’t think so.
ADI IGNATIUS: I wonder if Debbie Phelps told her kids, “Be gritty.”
ANGELA DUCKWORTH: I don’t think so.
ADI IGNATIUS: So we are out of time, but Angela thank you so much for your insights, thank you so much for being with us. That was fantastic.
ANGELA DUCKWORTH: Thank you, Adi. Thank you, HBR.
ADI IGNATIUS: That was Angela Duckworth, professor at the University of Pennsylvania and author of Grit, as well as the forthcoming book, Situated: Find the People and Places That Bring Out Your Best.
If you found this episode helpful, share it and rate us in Apple Podcasts, Spotify, or wherever you listen. If you want to help leaders move the world forward, please consider heading to hbr.org/subscribe to get access to the HBR mobile app, the weekly exclusive insider newsletter, and unlimited access to HBR online.
Thanks to our team, senior producer Mary Dooe, and senior production editor Kristin Murphy Romano. And thanks to you for listening to the HBR IdeaCast. We’ll be back with a new episode on Tuesday. I’m Adi Ignatius.
