ETL 35: Paul Michelman on AI’s Impact on Thought Leadership

Transcript: Paul Michelman and Bob Buday

Bob Buday: Paul, great to have you here. I have read so much about you and what you and Boston Consulting Group have been doing with generative AI. And I just knew I had to reach out to interview you and find out more behind the headlines and the speeches and demonstrations you’ve given.

Paul Michelman: Thank you for having me; it’s a delight to join you.

Bob: It feels like we’re we’re members of a small tribe. I think you saw that in New York City at the AI research services event. And we had that kind of tribe for six years in our thought leadership conferences. So but before we get into what BCG is doing with generative AI, I’d like to ask: How did you get into this thought leadership profession? And what attracted you to this profession, even though it probably wasn’t thought leadership when you entered it?

Paul: First, I think thought leadership is actually a relatively new term, at least relative to the amount of time we have been doing this work. I got into the profession, I suspect, in a way that’s not too dissimilar from a lot of other people, I fell into it sideways.

From the time I was 13-14 years old, I had a vision of myself as a journalist, and more specifically as a sports journalist. I followed that path as a high school student, into college, through to graduate school. I graduated from journalism school, at the very beginning of the downward spiral of the newspaper industry.

We were pre-internet, but the kind of fraying of the traditional display advertising model was already beginning and newspaper belts were tightening. After journalism school I came back to Boston, probably not the greatest strategy in the world at a time when its two primary newspapers were both laying off reporters.

That reduced my possibility of working there, which is not something you landed coming directly out of journalism school, not unless you’re extraordinary, and I wasn’t. But the entire market was flooded with journalists. I faced a choice of either going to a small town in the Midwest and to cut my teeth the way lots of my classmates decided to do, or look for another path.

I ended up at a job with a with Boston University, writing and editing for the admissions office. That led me on a path doing academic communications that eventually opened up a door. I was editing a newsletter for Harvard Business Review, which then kind of put me on a path of thought leadership and one that I have found immensely rewarding, but never intended to follow.

Bob: It’s kind of an accidental profession. I don’t know if you know, but I began as a sportswriter, too. Then I realized I was good but good sports writers are a dime a dozen. And I wasn’t going to be a great sportswriter. So you went through the academic communications route to discover this profession of thought leadership and you were at Harvard Business Review, the best-read and most prestigious management publication in the world. What did you do there?

Digital Transformation at HBR

Paul: I was extraordinarily lucky to be the right person in the right circumstance at the right time, with the right mindset. I joined HBR just after the turn of the century, and certainly internet, the web had been well-established media platforms by then — but not well established at Harvard Business Review. And in fact, I was hired to help convert a very old school, print newsletter into a digital newsletter.

I think I can lay some claim to being the first digital hire at Harvard Business Review. It’s interesting, you can’t be a one-person digital shop. I spent a couple of years working on a newsletter in print, dreaming of digital. And then there was some new leadership, including our CEO at the time. He recognized it was time for a transition to being a more digitally savvy organization.

He began to bring in new senior leaders who opened up the pathway to go digital. I ran toward that path — online work on the Internet. Working with digital products was already in my blood. I wrote scripts for CD ROMs. I worked on very early web pages for organizations. I was ready for the game.

But the appetite for the organization just didn’t grow quite as quickly as I wanted it to. There was a very clumsy period where we had built a digital team and I was no longer alone, I was now a member of 6, 8, or 10-person team who were brought on to lead, or at least ignite, this transformation.

But there was an awkward period. This will be familiar to lots of people who’ve experienced an old guard, new guard, and a lot of tension. And one senior editorial person for Harvard Business Review at the time, famously in my mind, said, “Harvard Business Review is to be blogged about; we do not blog.” That captured the mindset at the time. We tried to do this magic trick where we created something called Harvard Business Digital, which was a consumer-facing media offering, while maintaining the sanctity of the magazine.

The spark, the single most important moment in my career, came when somebody walked by my desk rather casually and said, “Paul, you heard about podcasting, right?” This was 2005. I had no idea what that person was talking about. So I looked into it and thought, “Wow, this is really interesting for us,” and launched Harvard Business Review’s first podcast, the HBR IdeaCast. It is still going very, very strong, thanks to the work of people who took this from me and built it. And that began to open up the idea of digital products for Harvard Business Review and the idea of creating new things that fit the emerging media that consumers were beginning to adopt.

The first HBR IdeaCast podcast came out [a year before] the first iPhone. We had apps, and there was the iPad, and you know, and then we had the social web opening up. So there was the series, in hindsight, of major technological innovations. We weren’t sure which ones to bet on. We also bet on some things that didn’t pan out. That was part of the game you played. But I landed in this space where I had the curiosity to go exploring, and was eventually given a good amount of latitude by some senior stakeholders at Harvard Business Review, to whom I am eternally grateful, to build out a new products function.

Continued Transformation at PWC and MIT

Bob: And then you went to strategy + business (at PWC) and you were there for about a year and a half and helped its transition from a print, newsstand publication magazine with a web edition. And you helped turbocharge the web version, right?

Paul: I was attracted to this position by nature of the opportunity to help do a transformation. Art Kleiner and his team had already made serious inroads, but I did get to help turbocharge that and turn us from a very much print-first into, if not truly digital first, digital equal. We started developing intentionally for the web, as opposed to repurposing for the web.

Bob:  I had Tom Fleming on the previous episode. And he said that the print product is gone and it’s all digital now.

Paul: They did that about a year and a half or two years ago. I thought that was a really nice move.

Bob: And then you went to MIT Sloan Management Review for about five and a half years as editor in chief and a podcast host. They had a digital edition, as I remember it, but you’ve kind of made them more digital, right?

Paul: I would say that Sloan Management Review was in some respects just a little bit more advanced than where HBR was a couple of years into my tenure. Culturally, it was very different a team, and actually quite eager to continue to lean in, on digital.

But for any number of reasons they really hadn’t made that big investment or big bold move to recenter, from print and print business model to digital and digital business model. That was a big part of the appeal to me. I think one of the reasons I was considered a viable candidate was my track record in digital transformation. I was lucky to be able to bring in some digitally savvy folks, but there was also an existing team that was ready to go. Within a year we moved from a quarterly release of content onto the web, synced with the publication of the magazine, to publishing something new every day. And that was a big transformation.

Personal ‘Multipliers’

Bob: When you look back at your career and when you got the digital bug, are there any people who had an outsized impact on you? I don’t know if you’ve read the book by Liz Wiseman called ”Multipliers.” It’s about these unique people in organizations who have a multiplicative effect on those around them — a positive effect — versus the diminishers or detractors, those who destroy morale and wind up with people who want to do less because they’re so they’re so disgruntled with the workplace. Were there any multipliers who made you realize you were fortunate to work with them?

Paul: I’m afraid I’m going to forget to mention somebody, but a few people have had an impact on me at very different times of career in my career. My boss when I was working for the Harvard Business School publications office was woman named Dorothea Bonds. And she was the first person who opened the door for me to go innovate. There’s a time in your career when you’re beginning to have the self-confidence to want to push the envelope. She really opened my eyes to the idea that the world is out there for your taking; you have to go define it; and she gave me space to do that. It didn’t necessarily result in anything like a great product breakthrough, although we did do some really innovative and interesting things. But it was a mindset.

At HBR, which was the formative experience for me professionally, I was blessed to work with outstanding colleagues. I’ve been blessed to work with outstanding colleagues every step of my career.

But that the HBR digital crew was a special group. And there were two people who me helped unlock my ability to produce value. One was woman named Jane Heifetz, who had been a longtime HBR staffer. She was an innovator at heart, and she unlocked that next level for me. We were in this place of deep exploration; we didn’t know what we were doing. And she said, “Just go do it, go do it.” That left a real mark on me.

The other HBR person was Joshua Macht, who was the publisher of HBR for many years and who now runs MassLive, the really excellent Massachusetts news, sports and entertainment site. He was a major enabler because he unlocked resources for me. Josh was the signature digital hire for HBR after being at Time magazine’s website, time.com. He was the big hire to come in and start blazing the path for the digital business. He understood me and what I was trying to do, and gave me the means to start building the business, not just play around. That was as exciting to me as the pure innovation. I love pure innovation. I love creating and working with teams to create new things. But when you can create a business or add to a business with those new things, that’s the home run.

Bob: I think this is so important for young people in their 20s or 30s, entering this profession of thought leadership, to realize that digital technology will redefine publishing, even more than what we’ve seen so far. And not just with generative AI but other AI, too. They need to make sure they’re joining an organization where colleagues and especially their boss are multipliers, using Liz Wiseman’s term. I’ve been fortunate to have those kinds of bosses, and I’ve been unfortunate to not have those kind of bosses. So how do younger people seek these people out? How do they know when they’re in a job interview that this person is going to be a multiplier, or the opposite?

Paul: I love that question. And I don’t have a good answer. I’m thinking back to my own experiences being an interviewee. Until maybe the last job or two I had, I wasn’t even processing that idea.

I was assessing the opportunity on really conventional grounds: What is it that this organization or this part of the organization needs? What can I bring to this that will I find challenging and exciting? Does  this place have an appetite for innovation? I think it depends on who you are and what you’re looking for. I would probably tell myself to ask questions that got to bosses’ tolerance for innovation, for risk and for failure. And maybe asking them about some of their own experiences in taking risks and failing.

Adjusting From Print to Digital

Bob: So, in 2016, you joined the MIT Sloan Management Review. I’ve worked with them since ‘87 when Sarah Cliffe was the editor in chief. In 2016, Bob Holland, who was the managing director back then, in a press release called you someone who has pushed the boundaries of publishing from the explosion of the web until now. What have been your biggest challenges for the print era people that you’ve worked with in pulling them into the age of digital publishing?

Paul: Maybe the biggest challenge is I don’t think there’s a single formula that works for everybody. First of all, I’ve had as many failures as I’ve had successes. And in part, maybe that’s because I hadn’t learned what I said earlier: that there isn’t [just] one way to do it; that you can assume that the things that motivate you motivate somebody else. Depending on who you are, depending on the length of time you’ve been in this world, depending on the specific job you do and how you do it, you’re going to be more or less wedded to your legacy ways of working.

The thing that’s most challenging is to change people’s mindset. I don’t think anyone is hopeless. By the way, I want to be very clear about that. I think the challenge is greater for some than for others. It begins with attitude — How do you coax the right attitude out of people?

For some people, it’s offering them the opportunity to participate in a project for a medium that is different from the one they traditionally work in. While we’re talking about the transition from print to digital, we could just as easily be talking about the transition from text to not text. Writing text for a long-form article that’s going to appear on a website is not that different from writing text for a long-form article that’s going to go into print. At the margins, it’s different. But the core activity is very different  — from writing a narrative arc for a podcast, to writing a script for a video, and writing for social [media]. We’re not even going talk about AI yet.

One thing I’ve tried to do is to open up opportunities to dip my toe in the water. [It’s the manager who says] “We need some help working on this podcast, and because it’s on a topic you know a lot about, we’ve come to you.” We did something like that with one of our longtime staff writers at BCG. That became an absolute mainstay, indispensable, for podcasts.

So it’s looking for safe ways to enter a new zone where you mitigate the threat. I like to be open with people when I see something that marks a major transition, but not to do it in a way that says, “If you don’t change the way you’re doing this work tomorrow, we’re all doomed.”

What’s the cliché: That we overestimate the impact of a new technology in the short term and underestimate it in the long term? When we start talking about AI, we might be experiencing some of that.

Overcoming Resistance to Change

Paul: There’s nothing wrong with people who want to continue to work on print. But you also have to give them signals that, “OK, we’re now beginning to go this way. It’s time to start moving over here.” I thought the task was going to be impossible at Harvard Business Review. Some people may have self-selected out along that path from “print is central” to “print as a complement to digital,” which is the state things are in now.

Bob: That task was difficult because they had a circulation of 300,000 plus for print at $100/year per subscription. That’s a lot of money. If you say, “Well, at some point, this is gone…”

Paul: That’s right. There were two primary motivators of resistance. And that motivation was found in senior business leaders who were unwilling to put this amazingly profitable operation at risk. And then there was the editorial issue, where senior editors at HBR didn’t really care about the business model unless it was beginning to threaten their livelihood; they cared about editorial excellence and thought leadership and their definition of how to create it.

That’s a different type of resistance. My world at that time was more about working alongside my editorial colleagues, bringing them into this new medium, than it was working to sell major business transformation. But I learned a lot by observing. I leveraged that when I landed at Sloan Management Review, where I worked alongside Bob [Holland] and others in making the case to the school to put us into an investment mode from a profitability mode.

It wasn’t that hard a case to make. I think they were ready for it because they believe in innovation. But I had learned a lot from that previous experience.

Bob: Many with a print mindset don’t even realize it’s in place. They think, “We can’t have too many articles, even if they’re all good.” That’s the mindset from the print era: There’s only so much space so even good articles would have to wait or be killed. This still happens with online publishing. Like, we have unlimited space…so why are we not publishing good stuff?

Paul: Another antiquated notion is connectedness between pieces of content. A magazine is a package; hopefully it’s carefully curated. A website is very different, and the experience is very different for people. They drop into an article, they leave, maybe you get them to stick around for one more. They’re not looking at the fullness of the experience. They’re there for the idea. That’s a big mindset shift.

Bob: I’ve also seen that even with articles they love; they don’t really want to know more about the author beyond a one-line bio. In a print world, I can understand why you don’t want to say too much about the authors. But in an online world, I imagine there are a lot of readers who want to know a lot more about the authors. And why don’t you give them the information?

Paul: I think that’s very true. If people are reading content from professional services firms, we of course have a vested interest in people knowing more about our authors. So we take a much more permissive approach to that aspect, for sure.

ChatGPT’s Early Impact

Bob: Open AI introduced Chat GPT on November 30, 2022. I remember that day. I imagine it caught your attention at the same time. Did the light bulb go off for you pretty quickly? Or did it take time?

Paul: For some reason that I cannot fathom, it took me a week to know about this. As I said earlier, it was a big moment for me was when someone walked by my desk and said, “Have you heard about podcasting?” and I had to go scramble and figure out what podcasting was. It was the same thing with ChatGPT. It might have been our chief communications officer who told me about it. It was several days after the announcement… it’s almost impossible to fathom how this happened, especially with my interest in in tech, but it just kind of zoomed by. And then I was in, and I was smoking. It was very much like the first time I opened the Netscape browser. I was aware of AI’s capability, but just like everyone else I’d never seen it. You never experienced it unless you were in tech, then suddenly, this tool was democratized.

Within a day, I was posting LinkedIn posts, written by ChatGPT. At first I was asking it to write limericks about BCG content, just doing all those silly things that we all did at the very beginning, because it was just so much fun to see what you could get this thing to do.

It took a while, relatively speaking, for me and my team at BCG to start figuring out: What is this thing to us? I think from the beginning, we saw that it is going to develop in a way that may start threatening the primacy of human beings and content creation. We’re still speculating about that.

Bob: Was there a fear that it would not just replace writers, people in thought leadership function, but also consultants – like, clients would just tap in their questions and they wouldn’t need BCG or McKinsey or Bain?

Paul: I can’t speak to that, because it’s not my part of the world. But I think it’s fair to say that very serious and deep conversations about what this technology would mean to the profession were ongoing before ChatGPT, and accelerated. I think across BCG, the real ethos is to lean into this technology, experiment with it, start building with it, and let’s chart our own course for how this affects us right across the organization. I began preaching that about six months in, when it was very clear that this was big. As time passed, as more of us started playing with ChatGPT, a really important moment for BCG was when we got a very early enterprise license with open AI, that every person in the company could access. That was a watershed moment, where anyone could go in and start writing their own GPTs.

We had that first zing of “Oh, my God, this is amazing,” for six to nine months, Then we were beginning to understand how we might be able to harness Gen AI for the longer term. And one was fairly obvious — as a tool of productivity, efficiency, and consistency. But an interesting happened for us. A second avenue opened up and kind of superseded the first one. And that was AI as a tool of experiential innovation. And what I mean by that is incorporating GenAI directly into content experiences, to do things that humans perhaps aren’t capable of doing on their own, and developing a real partnership between Gen AI and humans. Not just to create something faster, more efficiently, more consistently, but to create something that you couldn’t create on your own. And that’s made this journey a lot more fun and exciting for us, and a lot more rewarding so far.

Introducing GENE

Bob: So tell us about GENE, your conversational AI tool.

Paul: GENE is BCG’s conversational AI agent, which, for reasons that still kind of amaze us, has caught the attention of a sizable group of people out in the world, much to our delight but also surprise. And it really began with a small idea.  We were planning our podcast, which just completed its first season, called “Imagine This.” The idea of the podcast is to situate ourselves in a future state and imagine how something has evolved – for example, what does a C suite look like in 2030? It’s a way to help C suite leaders break out of short-term thinking and begin to entertain possibilities that aren’t obvious. And we thought that if we’re going to do a podcast about the future, we need a futuristic element as a part of the podcast. And we started brainstorming with elements that weren’t at all futuristic, like having a visual artist kind of draw the conversation on a board or something like that. Somebody suggested incorporating AI into the podcast. We all agreed that’s what we want to do, but we didn’t know what that meant.

So a member of our podcast team reached out to an engineer at BCG and asked him to come and think about it with us.  That opened up a conversation about conversational AI, and whether or not there was a viable way to bring kind of a live agent into a podcast, and we developed GENE from that. From a technology standpoint, it’s really fairly straightforward. There are some clever things that are happening, but it’s just layering of existing technologies. But the thing that made this work and I think continues to make GENE interesting is that kind of like serendipity. We had a team of podcast producers, writers, and engineers conceiving of an idea from the very, very start. Too frequently, new technologies are advanced so far by engineers that their purpose when they hit the market is overly constrained, and then eventually we all figure it out.

Bob: An old adage says that too much technology is a solution looking for a problem that doesn’t exist. What I’m hearing you say is, well, sometimes those technology solutions don’t solve an immediate problem, but they will solve eventual problem that people didn’t think of.

Paul: I’m being a little more gentle than technology as a solution in search of a problem. Sometimes, as you say, that’s a wonderful thing. Because we open our eyes and realize we can solve this problem that we weren’t even thinking about. What we actually wanted to do was develop this thing to solve a problem. We had we had the right group of people to do it, experts in these different domains. That led us to a very sophisticated prompt, and a very clever way of training GENE that has allowed us to produce this bot that has a personality… we really anthropomorphized AI. The voice is androgynous, and the name GENE is capital G-E-N-E, to try to keep focused on the fact that this is a technology, not a human substitute. But GENE’s magic is that it can have conversations with humans that are intelligent, and advance things in a way that humans may not have done on their own, because it’s able at once to follow conversations smartly and draw on its vast knowledge base. That’s really quite amazing.

GENE began with this one purpose, then we wondered else we could do with it and started kind of playing with other things. And then some of my marketing colleagues and media relations colleagues suggested we start bringing GENE to client events. GENE started getting media requests. It’s actually quite hilarious.

I had the privilege of appearing on stage with GENE at a thought leadership event a few weeks ago in New York. We’re continuing to evolve from just a podcast to other types of thought leadership experiences. This is a new way to deliver ideas to people. It’s a very different experience, and we could not do it without AI. And we’re going to continue to build GENE and the jobs it can do. GENE has gone from a single agent to a platform, which we call the genome. I didn’t come up with that, but I love it. It’s where we spin off different agents for different purposes. GENE is not constrained as humans are … there are now many GENEs out in the world serving different purposes. Interestingly, some of our clients are reaching out and interested in learning about how we’re using conversational AI. And that was not an original intent. But that’s a great outcome.

Bob: It’s fascinating to me. So GENE obviously can tap all the knowledge in its knowledge base, the web, and, and make connections that are not humanly possible.

Paul: Just like any large language model can. GENE is a deep expert in our ideas. It’s kind of a shallow expert with general knowledge, but its breadth is remarkable. And its ability to contextualize an idea in the broader landscape is very remarkable.

Bob: Do you think GENE will ever be capable of creating frameworks? I mean the Michael Porter value chain things…unique frameworks that take in all this information and say, “Well, here’s the two by two to think about this problem. Or the three by three.”

Paul: Let me re-phrase that: Would a custom GPT be able to do that? GENE would not be likely to return something like that, because it’s been trained for a very particular purpose. But could a product be built on a large language model to do that? I think it could be an instrument in doing it. But I believe there’s a creativity that humans bring to bear that I believe is still far, far superior than anything I’d have seen a GPT do. And a breakthrough framework — something that captures something the way, Porter, Christensen and Drucker did? I don’t think so. Now, I know that’s an answer today. That’s not an answer forever.

The Next Frontier: AI for Content Development

Bob: I remember [Michael] Hammer’s business systems diamond from our reengineering days. Can you imagine if a GPT can create pretty good frameworks that hold water within minutes, rather than the weeks or months that it takes super smart humans to do?

Paul: I think we’re going to learn a lot about that at BCG in the coming year, as we work on kind of our second project, a platform called Scribe, which really is kind of leaning into that opportunity around productivity, efficiency, and consistency. Scribe is kind of an end-to-end content development platform, customized for the type of work that BCG does, although potentially it could be customized for other work as well. That goes from the point of idea inception. Our teams of authors, who tend to be our partners in their cases, have learned something important by some combination of client work and research. And we want to turn that into a piece of classic consulting firm thought leadership. We are building out a tool that will begin with a GENE-like experience, where our authors have all their raw materials, their decks and data sets, and then have a structured conversation with a bot, which then builds out an outline or maybe something more like a story plan. And so that’s a type of framework, right? But we’re going to be learning a lot about AI’s ability to take a wide variety of inputs on a topic and make logical sense of them, hopefully in a way that would be well structured for the audience we’re trying to reach.

We are fairly well into developing that. We have work to do, but there’s real value to be created there.  That’s the first thing we took on and the most exciting thing. I’m sure this will resonate with lots of people: the early stages of creating thought leadership is endless churn and spin, and it’s painful and time consuming. And what we’re really trying to do is to have a lot of that happen between the principal architect of the idea and AI, cleaning this up together. Our AI agents are kind of challenging, coaching, and urging the person whose idea has the IP in it to a place where there’s a well-structured plan for a story, and a well-structured narrative properly supported with evidence. Bizarrely, I’m betting that the author is going to be more respectful with the AI than they sometimes are of another human being who’s challenging them. It won’t be personal for one thing.

Bob: So in other words, the outline or whatever the output is will look smart. But you really got to look at it closely to make sure the logic is there.

Paul: You’re absolutely right. For now and the foreseeable future, we expect that human beings – such as the writer who’s going to work on a project — will challenge and push. They may look at something and go, nope, this actually isn’t the best story. But we’re going to work from something that’s clean and logical, even if it’s not perfect. AI is really good at that, but we are the right sparks for creativity. We’re not interested in totally self-generating content. But AI can serve as an enabler and assistant to a human writer or editor. And it can accelerate the creation of ancillaries, offshoots, translations, localization, social media output…all of those things, and we’re going to build this out over the next year. It’s a pretty exciting prospect. And, you know, I don’t think it threatens human beings, but it threatens the way human beings do their jobs.

Bob: It also increases the quantity and quality of content that can be generated. I could see the whole thing being done by a GPT. You can feed it all the material from your consultants, lawyers, accountants or other professionals, and get a structure or an outline. The human professionals might say, “this is pretty good, this is wrong.” They might ask ChatGPT to find better examples. Then they could tell GPT to produce a 4000-word article as the Wall Street Journal or Harvard Business Review might write it.

Paul: If we if we did that today, I don’t think the quality would be there. First of all, we wouldn’t trust it, although hallucinations are becoming much less frequent. Contextualization mistakes are not infrequent.

I don’t disagree with you that that the future state is a possibility, even a probability. What I CAN say is that humans will play critical roles right in this process. I don’t think Gen AI threatens the role of humans in content creation in thought leadership, or in any form of content creation, but I think it will profoundly change our roles, and that’s an exciting and scary prospect.

How Humans Must Adjust

Bob: So is it too early to know which skills writers, editors, graphic artists, framework makers, and others must get much better at or develop from scratch? What would you advise your son or daughter in their 20s who love the thought leadership profession? They don’t want to go into a profession where they become the equivalent of the telephone operators, the switchboard operators, and be automated out of a job by the time they graduate.

Paul: I have a 25-year-old daughter who works in science communications and is thinking about going to journalism school, so this is deeply personal and relevant to me. I have a few thoughts, and they may not seem consistent. One, great writing will always be valued. Becoming a great writer is the cornerstone to becoming a great multi-platform communicator. You also need to develop proficiency in telling stories across different media. I think being a story orchestrator is a skill that will be highly prized. That means being able to plan a story across media types, understanding where you will use AI. What is this story in print? What is the story in audio? What is the story in AI-generated video? You need to be able to orchestrate the entire thing. In some respects, it’s kind of an executive producer role.

You also want to be able to work in one platform one day and another platform another day. You need to really understand the fundamentals of how to tell a story to an audience, and you should be able to tell lots of different stories to lots of different audiences. It’s flexibility. So many people in this profession, especially those of us who came up from the editorial side, began as a writer and expanded from there. But if I wasn’t capable as a writer, none of the other doors would have ever opened up for me.

For great journalists and great thought leadership writers, the bar is going to be ever higher. Because if AI can do your work just as well, then that’s where the work will go. I believe that really gifted storytellers, really talented writers will be more prized than ever before. I’m not sure they will be employed doing that job at the same volume they are today. So then, what does that mean? One is humans must learn how to work alongside AI. So we’re going to learn a lot at BCG, as we develop our Scribe platform, and we try to find that right balance of an AI assistant that lets writers make their own decisions.

Our appetite to open this up to people who aren’t writers will probably grow, and then we’ll ask AI to do more heavy lifting. But demonstrating the ability to work alongside AI — as a writer, editor, audio or video producer – has no downside. And I think it will be an expected capability pretty soon. Curiosity and willingness will help. You will be expected to be able to be a good prompt writer; that’s going to be like speaking English. Even if you have no professional interest in becoming a professional prompt writer, get in there and do it, because that’s how you understand what the technology is capable of. And hopefully, that sparks your own kind of curiosity to want to learn more, and lean forward into this.

For something like AI — which really is transformative the way the internet was, if not more so — you can drive or you can be a passenger. If you’re driving, you’re still on somebody else’s roads, and you don’t have infinite choice where you’re going, but you at least have some control over the destination and how you get there. If you’re a passenger, you’re at the whim of somebody else. And I don’t think you’re going to like the results nearly as much.

Bob: In the words of Alan Kay, who worked at Xerox PARC and was responsible for many of the inventions that showed up in the first personal computers: The best way to predict the future is to invent it.

Paul: Oh, amen, that’s beautiful.