Has AI Ended Thought Leadership?

By any conventional measure, I am a “thought leader.” I’ve spent decades at the intersection of innovation and storytelling, building companies, selling them, breaking things, and starting over. I’ve written six books. I contribute regularly to this publication. And I’m here to tell you the category is dying. Not because the ideas don’t matter. They do. But because we’ve entered a period where the ratio of people talking about the future of work to people building it has become grotesquely inverted. The content machine, supercharged by generative AI, has flooded every LinkedIn feed, every conference stage, every corporate retreat with an endless stream of polished, confident, and frequently hollow insight. We are drowning in it. It has never been easier to sound like an expert. A few well-prompted queries to an LLM, a slick carousel, a podcast appearance, and suddenly you’re a “future of work strategist” or an “AI transformation advisor.” The barrier to entry for expertise theater has dropped to zero. And when everyone can perform authority, authority itself loses meaning. I see this constantly in the work I have participated in at the Digital Data Design Institute (D^3) at Harvard Business School and with the organizations I advise. Leaders come to me frustrated. They’ve hired the keynote speakers and advisors. They’ve bought the frameworks. They’ve consumed the content. And yet their organizations remain stuck, paralyzed by the very insights that were supposed to unlock them. The problem isn’t a shortage of ideas. It’s a shortage of people willing to get their hands dirty testing those ideas in the real world. This is why I believe we’re witnessing the emergence of something far more valuable than thought leadership. I call it thought doership. What Thought Doership Actually Looks Like A thought leader tells you that AI will transform your workforce. A thought doer builds a pilot with your team in 10 days, figures out what breaks, and iterates. A thought leader publishes a framework for open-talent adoption. A thought doer restructures a business unit around a platform-based model and reports back on what happened—the ugly parts included. If that sounds like consulting, here’s the difference: A consultant delivers recommendations and then exits. A thought doer stays through the build, shares accountability for the outcome, and has skin in the game when things go sideways—which they always do. I’m an executive fellow at Harvard Business School—I believe in rigorous thinking. The distinction is about where the thinking happens. Thought leadership happens at a safe distance from reality, on stages, in op-eds, in beautifully designed slide decks. Thought doership happens in the mess. It happens when you’re running an experiment that might not work, with real money and real people, and you have to make decisions with incomplete information. Over the past two years, as I’ve helped organizations navigate the collision of AI and talent strategy, the leaders who have made actual progress share a common trait: They’re builders. Thought doership is a mindset, and it shows up both inside organizations, where leaders run real experiments with their own teams; and outside them, where the best operators-turned-advisors bring firsthand building experience to someone else’s problem. Thought doers don’t just consume ideas. They stress-test them. They prototype. They fail in small, instructive ways. And they share what they learn, inside their organizations and publicly, with a candor that most thought leaders wouldn’t risk, because it might damage their brand. The Faux-Expert Problem Generative AI has created a faux-expert crisis. I’ve watched people with no operating experience in a domain use AI to produce passable thought-leadership content on that domain within hours. The output reads well. It hits the right keywords. It refers to the right research. But it’s empty, because it wasn’t forged through experience. This isn’t just an annoyance. It’s a strategic risk for organizations. When a company hires a thought leader to give a keynote or facilitate a workshop, they’re often buying confidence, not competence. They’re purchasing someone’s ability to narrate the future, not their ability to navigate it. And there’s a world of difference between the two. So, how do you tell the difference? After years of watching organizations get burned and, frankly, after watching my own industry flood with performers, I’ve developed a pretty reliable nose for faux expertise. Here are the tells: The absence of scar tissue. Real operators have stories about what went wrong. They can tell you about the pilot that cratered in week three, the partnership that looked brilliant on paper and collapsed under its own complexity, the team that resisted a change initiative for reasons nobody anticipated. If someone’s narrative is all wins and frameworks, they’re performing, not reporting. “Altitude lock.” Faux experts are comfortable at 30,000 feet—the macro trends, the big shifts, the sweeping predictions. Ask them to drop to ground level, though, and they falter. What vendor did you use? How did you handle the procurement process? What happened when the union pushed back? What did the P&L look like in month four? Operators can move fluidly between altitude levels because they’ve lived at all of them. Performers can’t, because their knowledge was assembled from other people’s summaries. Generalities and platitudes concerning failure. Anyone can say “we learned a lot from our mistakes.” That’s content. A thought doer will tell you: “We assumed our internal team could manage the freelance marketplace integration alongside their day jobs. By week six, response times had tripled, the hiring managers were routing around the platform entirely, and we had to bring in a dedicated ops person we hadn’t budgeted for.” That’s knowledge. The difference is granularity, which can’t be faked. Rapidly accrued expertise. The faux-expert pipeline has a very specific shape: Someone reads about a trend, produces content about it, gets engagement on that content, and then starts advising on it, all within a matter of months, with no operating experience in between. Real expertise accrues slowly. If someone’s thought leadership on a topic predates any plausible period of hands-on work in that domain, that should be a red flag. Too much consistency. Operators update their mental models constantly because reality keeps forcing them to. Performers rarely revise their positions because their authority depends on the appearance of prescience. If someone’s point of view hasn’t evolved over two years in the fastest-moving technology landscape in human history, they’re not learning from experience. They’re protecting a brand. What to Do Before You Hire The smartest organizations I work with have started making a critical shift: They’ve stopped hiring thought leaders to inspire their teams and started hiring thought doers to experiment with them. But making that shift requires changing how you evaluate the people you bring in. Stop asking “What’s your perspective on X?” and start asking “What have you built around X, and what happened?” Any competent communicator can give you a perspective. You want someone who can walk you through the decisions they made under uncertainty, the tradeoffs they navigated, and the things they’d do differently next time. If they can’t do that with specifics, they’re the wrong person to bring in. Ask for the failed experiment, not the case study. I tell every CEO I work with the same thing: If the person you’re considering can’t give you a detailed account of something that didn’t work, they haven’t done enough real work to be useful to you. Case studies are retrospective narratives cleaned up for public consumption. You need someone who can tell you what the mess looked like in real time. Restructure the engagement model entirely. Instead of a one-day offsite with a charismatic speaker, design an eight-week sprint where an experienced operator embeds with your team, helps them design an experiment, runs it alongside them, and stays around long enough to learn from the results. The deliverable shouldn’t be a deck or a report. It should be a prototype, a pilot, or a set of documented learnings from something your team actually attempted. You’re not hiring a speaker or a traditional advisor. You’re bringing in an operating partner—someone who’s in your Slack, attending your standups, and co-owning the experiment. And vet the bio against the body of work. It has become remarkably easy to construct an impressive-looking professional identity, obtain advisory board seats, make podcast appearances, and publish “frameworks” without ever having run anything. Look for operational roles, not just advisory ones. Look for companies that the person you’re talking to has built or led, not just companies they counseled. Look for outcomes they were accountable for, not just conversations they participated in. Why This Matters Now The timing of this shift isn’t accidental. We’re in a period of technological change so rapid that yesterday’s best practice is today’s legacy system. The half-life of strategic advice has never been shorter. In a world where AI capabilities evolve month to month, any “thought leadership” that isn’t grounded in recent, firsthand experimentation is already stale by the time it reaches your inbox. I’ve lived this. When my co-author Jin Paik and I were writing our book Open Talent, we didn’t just study the platforms and the trends from a distance. We built companies on these models. The insights in that book were earned—through contracts that went sideways, partnerships that didn’t scale, experiments that produced results nobody predicted. That’s the kind of knowledge that moves organizations forward. The companies leading the AI transformation today aren’t the ones that hired the best keynote speaker. They’re the ones that gave a small team permission to build something, break it, learn from it, and build again. They created space for thought doership as an organizational practice. If you’re a CEO or CHRO reading this, here’s my challenge: Look at your next leadership offsite or transformation initiative. Count how many hours are allocated to consumption, listening to speakers, absorbing frameworks, and reviewing decks versus how many hours are allocated to building something. If the ratio is worse than 50/50, you’re investing in theater. Make a point of looking for people who have built in the domain you’re trying to transform. Not people who have written about it or advised on it, but people who have done it. Who have the scar tissue and the failed prototypes to prove it. Bring them in not to speak but to work. Make the deliverable an experiment, not an insight. To my fellow thought leaders—the writers, the speakers, the advisors, myself very much included—I’ll say this: The era where a well-crafted argument was sufficient to justify our value is ending. Our audiences are more overwhelmed and more burned out than ever. They’ve sat through the keynotes that didn’t translate. They’ve bought the frameworks that gathered dust. If we want to remain relevant, we need to be in the arena. We need to be building alongside the people we advise. We need to show our work. The future won’t belong to the people who describe it best. It will belong to the people who build it first.

Has AI Ended Thought Leadership?

Summary.

Partner Center