Reimagining business school through AI

Note: Transcripts are automatically generated from episode audio, and are not fully corrected for spelling, grammar, and formatting.

DAVID ROGIER: Most of us were taught in school and advised in our careers to pick an area and focus on it. I think that’s becoming a flaw for most people now. If you’re a copywriter in the top 1%, and you use AI, you can write even more world-class lines. But if you’re just an OK copywriter, a generalist using ChatGPT or Claude can do as good a job as you. What we’re seeing is a bit of a barbell effect, where the middle is getting thinner and you have the extremes of the specialists and the generalists. So what we’re doing is saying, “Hey, you’ve got to learn AI skills.” The second thing we said is, “How could we teach people to be more of a generalist?”

RANA EL KALIOUBY: That’s David Rogier, co-founder and CEO of MasterClass. For the past decade, his company has offered classes taught by some of the world’s greatest experts. We’re talking about people like Serena Williams and Hans Zimmer. MasterClass’ mission is that everyone deserves access to the world’s best teachers. David’s betting on AI to accelerate that mission even further. On this episode, we look at how AI is reshaping education, the AI tools that David uses every day, and how, amidst sweeping layoffs, you can future-proof your career. I’m Rana el Kaliouby, and this is Pioneers of AI, a podcast taking you behind the scenes of the AI revolution.

[THEME MUSIC]

Hi, David. Welcome to Pioneers of AI. It’s so great to have you on the show.

ROGIER: Thank you. I’m very excited to be here.

EL KALIOUBY: I heard that you just had a baby. Congratulations.

ROGIER: Thank you. She’s almost 1 year old, and she’s amazing.

EL KALIOUBY: We are going to talk about learning and education, but I thought I would ask: What is your biggest learning so far from being a parent?

ROGIER: Oh, God. It’s only been a year, so I am not an expert at all. My biggest learning so far is that every time she sleeps and wakes up, it’s like she got a software upgrade. She won’t be able to point before the nap, and after the nap, all she does is point.

Copy LinkWhy lifelong learning becomes a personal and professional superpower

EL KALIOUBY: OK, that’s awesome. Your daughter is a little too young for this, but with my kids, I have them sit down once a year and we do goal-setting. As part of that, we get very clear on our core values as a family. One of our core values is lifelong learning, which I believe is a shared core value for you as well.

Tell me more about that.

ROGIER: I was raised in part by my grandmother, and my grandmother lost everything in the Second World War. What she instilled in me was that the only thing somebody can’t take from you is your education.

EL KALIOUBY: I love that.

ROGIER: I think that shaped a lot of my life. I would ask questions in school all the time, to the point that my teachers did not like it. I was kicked out of the classroom. I was told to stop.

EL KALIOUBY: Oh, wow.

ROGIER: I was teased by other kids for it. I think it’s a trait that isn’t rewarded in education until much later in life, but I actually think it’s a key ingredient to being successful in life. I also think it’s one of the greatest highs in life. When you learn something that’s totally new and it changes how you see things, that’s like your own software upgrade.

Copy LinkHow a family story of resilience inspired the mission behind MasterClass

EL KALIOUBY: Yeah, that is true. You’re committing to yourself that you’re always going to be on this journey of getting software upgrades. I love that. How did your story with your grandma and your upbringing lead you to starting MasterClass?

ROGIER: It’s a great question. After grad school, I went to work for a professor of mine who ran a venture capital firm in the Bay Area. So I was investing. I don’t think I was particularly great at it. After a while, I went to my boss and said, “Hey, thanks so much. I really appreciate it, but I want to go build something.” He said, “What?” And I said, “I don’t know.” Then he wrote me a check for half a million dollars and told me to go think of an idea.

EL KALIOUBY: Wow, that is a great professor.

ROGIER: I was shocked. It was such a generous gesture. I know this is going to sound weird, but afterward it became a very tough time in my life because, for the first time in my life, I didn’t have any constraints, and I felt an immense amount of pressure to do something great. This man had trusted me with a lot of money, and you feel alone. You can’t complain about it to anybody because what’s the complaint going to be? “I have half a million dollars. I can do anything I want.” But you aren’t getting any hugs.

EL KALIOUBY: Right.

ROGIER: What helped me then was that somebody gave me a great constraint. She said, “Choose something that, even if it fails, you would be proud of.” That really helped.

EL KALIOUBY: That’s a great constraint.

ROGIER: Yeah. I sometimes think you need a constraint to be creative, or that it can help spur creativity. I mentioned that I was raised by my grandmother, and I remember once, I think it was in second grade, I went to her house after school and she sat me down. I was complaining about all the math homework I had, which obviously wasn’t a lot because I was in second grade, so I was just being a pain in the butt. My grandma told me she had a story she wanted to tell me, which, as a second grader, is the last thing you want to hear.

My grandma told me that when she was 16, she was living in Poland. She and her mom went on a family vacation. Her dad was going to join them, but he stayed home a couple of extra days to finish work. While they were on vacation, the Nazis invaded, and they took everything and killed her father. Her mom was able to flee through Portugal to Manhattan. The only job they were able to get was on a factory floor. They were working side by side, and my grandma decided she wanted to become a doctor while working in the factory.

She applied to every medical school in the state of New York. This was before computers, so she did it all by hand. She got a no from every single medical school in the state of New York, kept working in the factory, and applied again the next year. Again, she got a no from every medical school in the state of New York. Then my grandmother decided she was going to call the deans of admissions and ask, “Why am I not getting in?” I have chutzpah, but not that much chutzpah. They all hung up on her except for one, who said, “I’ll be honest with you, you have three strikes against you. You are a woman, you’re a foreigner, and you’re Jewish,” and then hung up the phone.

EL KALIOUBY: Oh, my God.

ROGIER: The fact that that person thought that was OK to say. Then my grandma kept working in the factory, applied again the next year to every single medical school, got into one, and became a doctor. I remember just staring at her because I had been complaining. All of a sudden, I got this intense lesson, and that’s when my grandma made it very clear: “Hey, David, this is the only thing that someone can’t take away from you.”

And I stopped complaining about my math homework. Going back to that moment when I had this one chance, I decided I wanted to try to build something that people can’t take from others. That’s how I realized it was going to be education.

EL KALIOUBY: First of all, it’s been a decade, right?

ROGIER: We just hit a decade.

EL KALIOUBY: Yeah, that’s amazing. Congratulations.

ROGIER: Thank you.

EL KALIOUBY: And I want to say the professor who gave you that first check is happy. Is he happy?

ROGIER: I mean, that ended up being financially a very good investment.

EL KALIOUBY: Okay. Well, a while ago, I wanted to take a class. I give a lot of keynotes and whatnot, and I really believe that I’m not that funny, and I wanted to really upgrade my sense of humor.

ROGIER: I’m sure you are.

EL KALIOUBY: Upgrade my sense of humor.

ROGIER: Okay. I mean, the Kevin Hart class and the Steve Martin class are great. I think the Steve Martin one you’ll really like.

EL KALIOUBY: I love Steve Martin. Okay.

ROGIER: The Steve Martin one, I think you’ll really like. Steve Martin basically says, “I’m not funny. I learned to be funny.”

EL KALIOUBY: Interesting.

ROGIER: Steve Martin’s approach to it is very close to what we do in product and tech. He basically tests the jokes out, and it was like 80% of the jokes. I think what’s so neat about that is it just shows what the craft is. Then, from doing that, you start getting intuition about things. But I think Steve Martin’s argument would be anybody can be as funny as him with just practice.

Copy LinkHow AI tutors can accelerate learning

EL KALIOUBY: Okay, cool. All right. I need to dig in and do this. All right, so let’s talk about how AI is shaping education. I guess one of the most exciting things about AI is that it does hold this promise of an idea that you have access to your own personalized learning tutor, right?

ROGIER: Yes.

EL KALIOUBY: You can ask any LLM of your choice any question that you’re curious about. It’s patient. It’s there for you all the time. Is there a difference between learning from ChatGPT versus, say, taking a class with Steve Martin?

ROGIER: The answer is not as simple as I would like it to be. The short answer is that with a well-trained LLM, you are able to learn the same amount as taking a class in a classroom, but in two-thirds less time. However, LLMs out of the box are not designed to teach you well.

EL KALIOUBY: Say more about that.

ROGIER: We ran a test on this. We took a sample of our students and ran three different groups. The first group watched a class. The second group used the study mode on ChatGPT. The third group used basically a combination of those, plus a pedagogy design to actually help you learn and stay engaged. That third group learned the same amount as the first two groups, but in 60% less time.

EL KALIOUBY: Interesting.

ROGIER: So what’s happening? LLMs out of the box are not trained in pedagogy and not trained on engagement. The trick we found, and how you should teach when you use an LLM, is that you want to front-load it with an architecture that causes you to think that you’ve learned. It’s not that you’ve actually learned; it’s that you think you’ve learned. Doing that upfront makes you more interested to learn. For example, Duolingo is great at this. You get points and a scorecard and all these things, and these goals you achieve. It’s not that you’ve actually learned that much, but that you feel you learned a lot.

EL KALIOUBY: Could I share about my Duolingo experience? Because I used to study French when I was in high school, and now I’m like, “It’s time for French again.” So I’m on my 85th day streak on Duolingo. I’m not going to break that streak learning French, right? So there’s something about that where it keeps you coming back. I don’t know if my French is any better, to be honest.

ROGIER: It’s probably not. There’s nothing really special about it, except it makes you think that you’ve learned, and it makes you excited to keep learning. That’s really important in the beginning of a process when you learn. Then we call it “hide the vegetables.” You want to sneak in pedagogical techniques that cause you to actually learn. If you do that right, it’s both engaging and you learn. It beats an LLM by far. So this is what I mean: Out of the box, it doesn’t have that effect.

But I’m going to tell you about a crazy story. The World Bank ran a study in Nigeria, I believe, with 1,000 schoolkids who were all learning English. Half the kids just learned in the classroom. The other half used a pedagogically guided version of ChatGPT a couple of times a week in the afternoons. What the World Bank found is that 18 hours of interacting with ChatGPT in that pedagogically guided way was the equivalent of a year and a half in the classroom.

EL KALIOUBY: Wow.

ROGIER: And the results held after a year. So for you, when you’re learning on Duolingo, if you did it in a pedagogically guided way in ChatGPT, with the same amount of effort, you would have learned probably 3x.

EL KALIOUBY: I’m not an expert in language learning, but I do wonder what platforms or approaches, especially boosted with AI, will actually help me feel confident speaking a full sentence in French at a party. We’re going to take a short break. In a minute: digital twins, and why you won’t find a video AI clone of Steve Martin on MasterClass yet. Stay with us.

[AD BREAK]

Copy LinkWhere digital twins add real value

Okay, so I want to talk about digital twins. Something I’ve been experimenting with personally is basically creating both a chatbot version of me and an actual video version of me, right?

ROGIER: And what are you using it for? Are you using it yourself, or are you using it for, say, when somebody wants to hire you to speak but they cannot afford you, so they hire that?

EL KALIOUBY: Yeah, I’m actually experimenting with that. I feel like my digital twin isn’t there yet because it doesn’t have my, I don’t know what I call it, my spirit. It doesn’t have my smile. It’s like buttoned-up Rana, which is not me. And then the second piece of it is there’s usually Q&A at the end of these presentations and keynotes, which is actually my favorite part. I don’t trust my digital twin to answer these questions on my behalf. But I’m curious: From a production standpoint, it would be much more cost-effective for you guys to have Steve Martin create a digital twin of himself in three minutes, and then use that to create, I don’t know, a six-month program or whatever. Have you explored that? What do you think of all of that?

ROGIER: We’ve thought about and explored parts of it. So, right now in beta, we allow you to interact with a digital twin, a voice twin, of approximately three instructors. One of them is Mark Cuban, so you’re able to pitch your ideas to him, and he’ll talk back and give you notes and things like that.

EL KALIOUBY: Amazing.

ROGIER: What we found is pretty magical. When he comes on the phone for the first time, you get anxious, like, “Oh my God, I’m talking to Mark.” There’s something magical about it. We’ve also trained it, with his permission, on his books and blog posts, so it’s a pretty good version of Mark. The other thing is, if you look at ChatGPT, it’s trying to go so broad. Because we’re only talking about entrepreneurship and business, you’re able to get it to pretty much stay within those bounds, too, and you get over the AI part pretty quickly. It’s interesting, the types of behaviors we have to adjust and change. I’ll give you an example. Engineers are going to hate that I’m sharing this. I was looking at one of the early versions, and it was great. Then I saw a later version, and there was a longer lag between when you finished talking and Mark started talking. I was like, “Why is it so slow now?” They said, “We had to put that lag in.”

EL KALIOUBY: Oh, really? To make it…

ROGIER: Because people got upset that it was interrupting them. However, on the video side, I haven’t found something I’m comfortable with yet, though I’m just starting to see ones that might be good enough. For us, I can find the use case I’m comfortable with on voice because you know it’s AI, we’re very clear about it, and it’s something you can’t do without AI. But on video, I don’t know. It feels like it ruins some of the authenticity of the person. So for now, we’ve been staying away from a video version of any of our instructors.

EL KALIOUBY: Yeah. I like the MasterClass On Call, right? The voice part, because…

ROGIER: Exactly.

EL KALIOUBY: …we had Mark Cuban on the show, and he’s awesome, right? He’s great, but he’s not scalable.

ROGIER: Exactly.

EL KALIOUBY: Not everybody’s going to get access to Mark Cuban and have a conversation about their idea. So what this does, I guess, is allow potential founders and entrepreneurs to pick his brain.

ROGIER: Exactly. I have a strong point of view on this. I think a lot of us are having the wrong conversation about AI in our companies and product offerings. Most of the conversation is, what can I replace with AI, or how can AI improve this? I think about this as the Martin Scorsese test. I’ll give you an example. I could use AI to make a film, but Martin Scorsese could make a better version of it. So that’s not where I should be spending time. I should be spending time on what I can build that Martin Scorsese could not do. For example, instead of making one great film, what Martin Scorsese can’t do is make 10,000 films that are personalized for every person. So what AI should be doing on the video and media side is creating 10,000 individual shows for everybody. In the same way in education, instead of just replacing a video version of Mark Cuban, which I could film and would be good, what I can’t do, to your point, is get him to coach every one of his students. But I can with AI.

Copy LinkHow AI changes jobs and why generalists may gain an edge

EL KALIOUBY: Right. My investment thesis in my fund is human-centric AI, which is AI that unlocks human potential and amplifies and augments what we do, not replaces us.

ROGIER: Okay. Can I ask you about that?

EL KALIOUBY: Yeah, ask me about that. We’re going to get spicy here.

ROGIER: Do you think it doesn’t actually replace people?

EL KALIOUBY: I think some jobs will go away, but I actually think that for a lot of jobs, it’s not just a combination of tasks that AI can replace. Yes, maybe it will replace specific jobs or tasks, but not many, because they are more than just automatable AI tasks. I also think that AI is going to create brand-new opportunities and brand-new jobs. We’re seeing those unfold in the same way that social media created a whole slew of new opportunities, like social media influencers and social media marketers and whatnot. So I think it’s going to be the same with AI. But what do you think?

ROGIER: I hope you’re right.

EL KALIOUBY: Okay.

ROGIER: I hope you’re right. I was looking into this around the advent of the spreadsheet because, if you remember, there was a whole industry of people called clerks who were doing basically what a spreadsheet does. Then the spreadsheet comes along and erases that, and all those people lose their jobs. That kind of displacement has major impacts on your income and even on your health. But it also led to the rise of the analyst, and there were tons more analysts. So in the net, it was a net gain. So, A, how do you help people who are in the position to lose? But I think the difference, potentially, with AI compared to your example of the internet, the PC, or the spreadsheet, is that there are two risks that make it different. One is the speed of adoption. One of my fears is, does the change happen so fast that we can’t catch up, and then it becomes a much bigger issue? Or two, is there something different about AI, because it’s replacing thought, that causes it to not behave the same way? But I do think there are going to be a lot of people who lose.

EL KALIOUBY: Yeah. I do believe there’s going to be a big transition in what jobs look like. Even with my 22-year-old daughter, I can already see that these entry-level positions are changing. They look very different. I agree with you that we’re not ready as a society. The technology is moving way faster than we are ready, as organizations or as a society, to absorb it.

ROGIER: Yeah. And we are seeing this at MasterClass internally, and we’re also seeing it externally. Some large employers came to us and said, “Hey, we are noticing some gigantic shifts in that people don’t have the skills that they need. The types of people we need are totally different.”

EL KALIOUBY: Yeah. What have you been doing to reskill people?

ROGIER: If most of us were taught in school and advised throughout our careers to pick an area and focus on it, I think that’s becoming a flaw for most people now, and a trap. If you’re in the top 1% of that skill set and you use AI, the AI is making you thrive. If you’re a copywriter and in the top 1%, and you now use AI, you can write even more world-class lines. But if you’re just an okay copywriter, a generalist using ChatGPT or Claude can do as good a job as you. So what we’re seeing is a little bit like the barbell effect, where that middle is getting thinner, and you have the extremes of the specialists and the generalists. The middle is getting thinned out.

So what we’re doing is saying, “Hey, you’ve got to learn AI skills.” We’re doing lots of classes in collaboration with NVIDIA, Microsoft and OpenAI to teach people the core skills they need, and those have been doing fantastic. The second thing we said is, how could we teach people to be more of a generalist? So we actually launched a reimagined business school-like experience with the University of Chicago Booth and OpenAI. It’s a 12-week program, and it teaches you not only a broader set of business skills, but AI within that broader set of business skills. I think the business rules are being rewritten, and we are trying to provide something not just for our employees, but for everyone in the world.

EL KALIOUBY: After a break, David shares more about this new reimagined business school, plus more on why we will still need management skills, even when AI enters the org chart.

[AD BREAK]

Copy LinkWhy business education is being rebuilt for an AI-native world

I do actually want to click on the MasterClass Executive, right, which is this partnership between OpenAI, the University of Chicago, and MasterClass. I’m curious, how did that come about as an idea?

ROGIER: It came out of a paternity leave. During paternity leave, one thing I did was read a bunch of academic articles on AI and education. Then we started talking to Bain, Deloitte, Dick’s Sporting Goods, folks at Lovable and at ElevenLabs, and they all said the skill set they need, they aren’t finding in the job market. So we thought, what if we could make something that people would actually really want?

EL KALIOUBY: What is this skill set, by the way?

ROGIER: It’s AI-native business skills. We reached out to Chicago Booth, which is a top-three program in the world, about it, and they have some cutting-edge research and work on it, and OpenAI obviously does as well. We said, “Why don’t we join forces and actually do it?” I think why it’s hitting such a nerve is that we’re afraid we’re going to become out of date. An interesting thing is that a lot of people applying, we thought it was going to be people looking at it as an MBA replacement. It’s actually executives who are applying for it, which is really fascinating.

EL KALIOUBY: That’s fascinating.

ROGIER: It’s a very impressive group of folks who are applying. I also think online education in the academic world has gotten a bad name because of a lot of the scammy institutions out there. So our idea was, “Hey, by combining MasterClass, OpenAI, and Chicago Booth, you’re combining the best of three different worlds, and it’s a stamp of approval. Not everybody gets in. It’s hard to get into.” So, can we create that new credential that people really care about?

EL KALIOUBY: Give me some examples of the classes that are in this program.

ROGIER: The stuff you learn is amazing. I got an MBA at Stanford, and it was a great experience, but the classes are no longer what you would expect in an MBA program. Some of that is still really important in this world, but we also have a class on managing an AI team. So, what happens when you have 100 agents and your job is to manage them? It is a management job. You have to give them feedback and notes. You have to set expectations. You have to decide which ones to keep and which ones to dump, how they interact with humans, and how you’re going to check the quality bar, all that.

EL KALIOUBY: I agree with you. Very little mindshare is being spent on how you build a culture in an organization that is hybrid human and AI.

ROGIER: Yes.

EL KALIOUBY: Accountability, trust, right? All of these things that are so important in a functioning company — how do you replicate that?

ROGIER: I think these rules are being written now, but we try to take the people who are the most cutting-edge on it and are building it and trying to figure it out. One of the big things that comes up from humans is, “Am I still doing my job if I do these tasks with AI?” I thought that was really interesting because I think managers aren’t great yet at being clear on what the expectations are. There’s an expectation that you learn AI and get really good at it, but can my entire job be done by AI? What if it’s all done by AI, but I stand behind every word of it, and I’ve read it, and I stand by it? Where does that line get drawn?

Copy LinkWhat it looks like to manage AI teammates alongside human ones

EL KALIOUBY: Yeah, that’s so fascinating. At my fund, we’re experimenting with an AI chief of staff.

ROGIER: Cool.

EL KALIOUBY: We’re using Anthropic’s Claude to do that. We initially did not give it a name, and then we were like, “You know what? We should give her a name.” Her name’s Blue. My fund’s called Blue Tulip Ventures. B-L-U-E is her name. The biggest task I’m personally using Blue for right now is if we get an incoming potential investor, LP, and/or founder, I want to log this in the CRM.

ROGIER: Yeah.

EL KALIOUBY: But it’s not just a matter of logging it. It’s actually a set of tasks. You have to go find them on LinkedIn and research them a little bit, find their email, and figure out who made the introduction if they came through a warm intro. So, it’s a whole bunch of five to 10 tasks that are stuck together. Every time this happens, I make a decision whether to send it to Blue or to send it to our junior team members.

ROGIER: And?

EL KALIOUBY: I feel like on the LP side, I have it down with Blue. It knows what to do. It’s pretty accurate.

ROGIER: Is Blue sending emails or just drafting them, or how far are you letting Blue go?

EL KALIOUBY: Not sending emails without a human in the loop.

ROGIER: Yeah.

EL KALIOUBY: I’m in the loop for sure, but I will have it draft emails. I will have it update the CRM.

ROGIER: Yeah.

EL KALIOUBY: I will have it nudge me to do things, but not on the founders, on the startups.

ROGIER: Why?

EL KALIOUBY: Because I don’t trust it yet. For example, we categorize every start-up that comes in. We categorize the stage of it, but also whether it’s in the health and wellness space. I don’t trust its categorization yet, but I wonder if I’ll trust it more over time, right?

ROGIER: I also use one.

EL KALIOUBY: You do?

ROGIER: Yeah, I mean, I have it draft my Slacks and my emails.

EL KALIOUBY: Interesting.

ROGIER: I have found some of the most impactful things are that it prioritizes my to-do list. So, it’s taking from my Slacks, from docs, and from emails, aggregating all my to-dos, and then, knowing what I’m working on and how much time I have, it creates a prioritization for me, which I find really helpful.

EL KALIOUBY: Cool. I’m going to do that.

ROGIER: Other things are just helpful and neat on the side. I realized one of the things I do is give notes on classes that we make. I found myself wondering if I actually give the same notes on a repetitive basis. So, I analyzed all my notes, and I created a little bot for our team that they can input the class into and get notes based on my previous notes. They still meet with me. I still talk to them, but it lets them get a head start on it.

EL KALIOUBY: And you’re not the bottleneck, too, right?

ROGIER: Exactly.

Copy LinkThe AI tools that a CEO uses daily

EL KALIOUBY: I love that. Actually, I saw that you shared on LinkedIn your CEO AI stack.

ROGIER: Yes.

EL KALIOUBY: So talk us through that. What’s on there? Then I’ll share what’s on mine.

ROGIER: Yeah, and it’s always changing.

EL KALIOUBY: Okay.

ROGIER: A couple of things. One is, I use an LLM every day for almost everything. At the bottom of the stack is ChatGPT, Gemini, and Claude. For different things, I like each of them a little differently. Then what gets more interesting is learning. I learn a lot now with NotebookLM. I put in a bunch of academic articles, and I listen to customized podcasts now.

EL KALIOUBY: By the way, that’s how my son studies for all his exams. So interesting.

ROGIER: I create a bunch of apps for myself in Lovable. Right now, I’m tracking what I eat, so I made an app in Lovable to do that. It takes a couple of minutes, and it’s customized exactly to how you want it to be. There’s no reason to buy an app. It takes five minutes to do it. I’ve created probably six of those apps on my phone that I use every day. Then, on the chief of staff side, I’ve played with different versions of it. I’ve done one where I enter all my to-dos every day. I have a to-do app, and then in Make.com, it sends them to ChatGPT. It does the prioritization, and then it puts them back into the to-do list.

Now I’ve tried it with the coworker side with Claude, which is taking more of my connectors and docs on there. One of the tools we built internally at MasterClass that we now have in test, so I use this all the time, is what we call MasterClass Coach for right now. It listens in on your Zoom calls and gives you notes both in real time and afterward on how you did in that meeting and what you could do to improve.

EL KALIOUBY: Oh.

ROGIER: I’m going to share a personal story from it. I did this the first time, and I was in a meeting with my chief of staff of the last 10 years, my human chief of staff. The notes told me that I’m not asking her enough questions. I was like, “No, it’s just because I’ve worked with her for a decade.” The next meeting I’m in is a three-person product meeting. It gives me the same note. I’m like, “No, it’s just because I’m the CEO. I’m telling the team what to do.”

EL KALIOUBY: This AI is biased. It’s hallucinating.

ROGIER: “This AI is biased. It just doesn’t know the context.” Then, the third meeting I’m in, it gives me the same feedback. I was like, “Oh, God, okay. I wonder if it’s me.” So for the next week, I decided, “Okay, I’m just not going to state my point of view so quickly. I’m going to ask more questions.” Unsolicited, a day and a half later, an employee texted me and said, “Hey, David, I just want to say, the way you appear in our meetings now, you ask a lot more questions. You’re much more engaged. I just want to say it’s really awesome.” And I was like—

EL KALIOUBY: Wow, that’s awesome.

ROGIER: “God damn it. That was right.” I found that it’s been an amazing tool. We used it internally, and now we’ve put it out in the world.

Copy LinkWhat it takes to thrive as AI keeps reshaping work and identity

EL KALIOUBY: That is so cool. I love that. Last question: What do you think it means to thrive in the age of AI?

ROGIER: I think thriving in the AI world means recognizing that you’re going to have to keep learning because everything changes every week. It’s going to require accepting that the views I had of my own value and worth and job are going to change. Three, it has to be an openness to the new and optimism about it. Fourth, it’s also a healthy dose of fear: “Hey, there are certain areas where we want to run as fast as we can, but there are certain areas where we probably want to go slower, or where we want to put some caps and controls on.” I think thriving means doing those four things.

EL KALIOUBY: Yeah, I love that. Well, thank you so much for joining me on the show. This was great.

ROGIER: Of course. I really enjoyed it.

EL KALIOUBY: That was my guest, David Rogier, co-founder and CEO of MasterClass. Before we end the show, I want to say goodbye and a big thank you to our producer, Rachel Ishikawa, who’s moved on to a new role. Rachel has been with the show since its launch back in 2024, and she’s been so instrumental in shaping it. Rachel, thank you so much again, and we wish you the best of luck.