Hill — The Leadership Skills for CEOs That Will Matter Most in the Age of AI
After spending decades advising CEOs and studying leadership, Harvard Business School professor and author Linda Hill has developed a clear view of what makes great leaders — and where many fall short. In this episode of Leaders with Francine Lacqua, Hill explains why today’s most successful executives prioritize collaboration over charisma, embrace constructive conflict and create organizations that can adapt to constant change. She also shares how artificial intelligence is reshaping leadership and the skills the next generation of executives will need to thrive.
TL;DR
A ~24-minute interview on Bloomberg Podcasts’ Leaders with Francine Lacqua series, published 16 August 2026. Linda A. Hill — professor of business administration at Harvard Business School, faculty chair of its Leadership Initiative, principal of the consultancy Paradox Strategies, and author of Becoming a Manager, Collective Genius and Genius at Scale — is interviewed by Francine Lacqua. Lacqua frames the episode as an inversion of the series’ usual format: rather than asking a leader how they led, she asks a career-long student of leadership “what actually makes a great boss?”
The episode’s most useful contribution to this wiki is §7 below — a named distinction for leading when no vision is available, arrived at in the context of AI specifically.
1. Self-knowledge as instrumentation, and the abdication of omniscience. “A successful leader really has to know himself or herself, because it really starts with you. You’re trying to use yourself as an instrument to get things done. And if you don’t know yourself, it’s hard to do that in a very deliberate way.” Then: “your job is not necessarily to have all the answers.”
2. Contextual intelligence — the construct she treats as under-supplied. “You need to have emotional intelligence, but you also need to have that contextual intelligence to be able to read things.” The operational content is an attribution-error correction: “usually our context is shaping our individual behavior. Too often we think, oh, that person’s a good person, bad person, resistant, whatever. No. What is it about their context that is leading them to prioritize in ways that aren’t suited to what the company needs? Or are they constrained in some way that if I relieve that constraint, they begin to behave in ways that the company really needs?”
3. Duties over privileges, and connection as the failure predictor. “I almost always start by saying, think about not the rights and privileges that come with being a CEO, but the duties and obligations. And people can tell if you care… because you’re going to make many controversial decisions, and you’re going to have to deal with trade-offs, and they want to know about your character.” The relational framing she flags as a deliberate update: “not so much as following you, but co-creating the future with you. And if you can’t get that connection going, that’s when you know there’s going to be a failure.”
4. Creative abrasion, and the Steve Jobs / Ed Catmull illustration. The organisational-capability question she says CEOs should be asking: “do you have enough of what’s referred to as creative abrasion? Do you have enough diversity of thought? And do you know how to embrace that diversity of thought, deal with that conflict? You need to amplify difference. But many CEOs minimize difference.” The framing: “everybody has a slice of genius… if you have very talented, passionate people and you let them say what they think, they’re not letting go easily. So you as a leader have to be able to take those talented, passionate people — if you’ve unleashed their ideas, now you have to leverage and harness them to make sure you get some good solution out of it.” Her Google infrastructure-era source: “you’re hiring them for their ideas, and if you don’t get those ideas, then shame on you” and “you know how hard it is to do business nowadays? You can never have too many cooks in the kitchen.”
The Pixar anecdote is the sharpest illustration of contextual intelligence in the episode, and it is about a leader removing himself: “one of the things that Steve Jobs understood is that most action that happens is collective action. There would be times when Steve being in the room was not going to be possible, because he was going to throw off the dynamic too much — not because he couldn’t behave well necessarily, but because other people would be uncomfortable with him, so he would remove himself.” Ed Catmull is credited with the read: “he understood people and what they needed and didn’t need, and what the dynamics would be of a group if a certain mix was there… If he thought it’s not the right mix, then he had no problem saying to you, Steve, you can’t come.”
5. The feedback vacuum at the top. “As you become more senior, people don’t give you feedback. That’s mistake number one. You need to be sure that you have an organization that tells you when problems crop up. And many, many senior people do not have organizations like that. They will tell you that — they will admit it.”
6. Portability of talent — the negative finding. Citing HBS colleagues Boris Groysberg and Nitin Nohria on predicting whether CEO-race runners-up succeed elsewhere: “what they have found is that people aren’t as portable as people think. And it’s often because… selecting on competence and not paying attention to what’s their network of relationships. People who are successful in one context, turns out you’ve been successful in part because of the network of people you’ve been working with who know you and in various ways help you — and when they move to the new environment, they don’t have that group, they actually struggle.”
7. Wayfinding vs. pathfinding — the episode’s most transferable idea, and it is about AI. She sets it up by separating leadership from management via Kotter and Bennis (“leadership is about dealing with change and management is about dealing with complexity — and we have complexity and change happening now”, with large firms tending to be “over managed and under led”), then distinguishes the innovation case from the change case: “innovation is not so much about vision. It’s more about purpose… vision means you know the direction. When you’re talking about innovation, by definition, if it’s a breakthrough innovation, you have no answer, you have no vision to tell people, this is it.”
Applied to AI, without hedging: “many leaders have no vision about what’s really going to go on with AI. They just don’t. When you ask them, they’re perplexed.” Lacqua offers the charitable read — “which is maybe honesty, Linda… at this juncture in time, maybe we don’t know” — and Hill accepts it and names the alternative competence: “so that’s why we call that wayfinding as opposed to pathfinding. The wayfinding is we don’t know the destination, nonetheless how we’re going to get there. That requires a lot of confidence in your capacity to help people figure this thing out.”
The worked case is the head of Cleveland Clinic Abu Dhabi — a Rhodes Scholar and robotic-cardiac-surgery pioneer — who told her “I feel like I’m leading through a fog. How do you lead when you can’t see, when you have no vision?… I thought I should steer the ship, but where am I steering it to?” His resolution, and the actionable part: “my job is first to convince everyone in this organization that there is nothing called business as usual. Everything is a working hypothesis.” Hill reports that the hospital handled Covid unusually well and attributes it to a specific staffing rule that follows from wayfinding: “every time when he could, he made sure that there was someone on the team making decisions who had never seen an epidemic before. He didn’t just want the experts, he wanted someone who had the beginner’s eye, who would ask that so-called stupid question — because they didn’t know it was a stupid question — and would force all of the other experienced people to question their assumptions.”
8. Decision rights as the CEO’s first design task. “One of the things that I first work with when I’m working with CEOs is decision-making rights. Being real clear who has decision-making rights for what, and what do you want to hear before they make a decision. And how do you need to set up the way you work so that you’re comfortable having delegated that.” The reason is a distributed-sensing argument: “you build an agile organization that can adapt to anything… because somebody down there is going to hear it sooner than you, and you want to make sure you’ve empowered them enough. There have to be guardrails here.”
9. The unspoken risk-appetite divergence — the episode’s cold open, and its most specific governance claim. “The other thing that very senior people don’t necessarily talk about — and even boards — they don’t talk about risk really. They can have very different risk appetites… but they don’t ever really address the fact that they have different risk appetites.” Asked whether this extends to board members: “no, because I don’t think board members like to have conflict. Conflict is complicated. So sometimes — do we bring it up or not? It’s there whether you bring it up or not, and it’s going to play out in different ways. But CEOs or board chairs or lead chairs understand that. Nowadays, given what’s going on, we need to talk. We need to say what is our risk appetite as we begin to, for instance, roll out AI solutions. And that conversation is a really hard one to have.” She adds the reason it stays unheld: “how do we have that so we don’t make that person look like — oh, you’re too conservative, or you’re resistant to the future.”
10. Charisma, deflated. “Charisma is a source of power. First off, we need to feel that it’s authentic.” Asked for the most charismatic leader she has met, she names Celtics head coach Joe Mazzulla and immediately re-describes charisma as deliberateness: “I was just blown away with how deliberate he was about everything he does.” The illustration is scope of relationship rather than presence: he built relationships not only with players but with kitchen staff, the business-operations side (“they were somewhat surprised at how much time he spent with the business operations, not just the basketball operations”) and the media (“he invited them in and was very transparent, because he understood that it was the whole”).
11. Co-CEOs. Common in tech as “two in the box” (technical visionary plus business operator), now appearing at the very top — Netflix, Oracle, Spotify, Comcast. Her conditions: “it does [work] if they get along very well together and if they can be honest with each other”, plus “a clear division of labour.” Two rationales she has seen: neither candidate individually covers where the business is going; or one is assigned the future and diversification while the other runs the core. The failure analogy: “you’ve been in companies where you’ve even had the CEO and the COO don’t really see eye to eye — the rest of the organization has trouble being aligned.” She notes the field’s own shift in unit of analysis: “when I first started as a professor, you really focused on what’s the CEO doing. And then by the late ’90s, you really wanted to understand how that senior team worked.”
12. Rapid-fire — one answer worth keeping. Asked which leadership practice is overrated: “Empowerment, as people think about it… Empowerment simply means that you’re sharing with people the potential to have influence. So whether or not they get to exercise that potential is your call. And I think sometimes people [say] ‘I empowered them.’ Well, in the end it’s on you. You need to make sure that they actually know what they need to know or they’re going in the right direction. Otherwise, you need to take back that potential and tell them, not this time.” Asked which skill will matter more in ten years: “your capacity to collaborate.” Asked what management advice she disagrees with: “that you need to make sure that people know you’re the boss. That feels very ’80s — but trust me, there are many people who want you to be clear about who the boss is. I think formal authority as a source of power is actually quite limited in what you need to do now to influence people to get the job done.”
What was actually ingested
The complete episode — all 389 caption segments across duration: 23:38 / length_seconds: 1418, with all six chapter markers. Captions are human-curated (the video carries an auto-generated track as well), so transcription quality is high; only two Harvard faculty names needed correction (Boris Kreuzberg → Boris Groysberg, Nithya Noria → Nitin Nohria).
Two structural notes for anyone quoting timestamps. The episode opens with a ~30-second cold-open pull-quote taken from the board-risk-appetite passage at roughly 20:30, before the “I’m Francine Lacqua” title card — so the chapter boundaries do not align with the first minute of speech, and that quote appears twice in the transcript. Second, Lacqua’s spoken introduction and the three inter-segment narrations are scripted voice-over, not interview.
A correction I made and then reverted, recorded because the reversal came from the wiki itself. An initial cleanup pass changed Lacqua’s “Genius at Scale” to “Collective Genius”, assuming a slip. That was wrong: the wiki’s own Ross & Schneider source already cites Genius at Scale as a forthcoming Hill title in February 2026. Lacqua’s list is correct as spoken and the raw file has been restored.
Scope warning. This is an interview, not research output. Hill draws on decades of fieldwork (Pixar, Google, the Celtics, Cleveland Clinic Abu Dhabi) and cites colleagues’ studies, but nothing in the episode is presented with a method, a sample or a result. The Groysberg–Nohria portability finding is the only named study and is described in one sentence. The AI content is confined to roughly the last eight minutes and is framed as observation about the leaders she advises, not as a finding. There are no numbers anywhere in the episode.
On AI specifically, note what she does not claim. She does not say AI changes what leadership is; she says the conditions have become ones where the pre-existing distinction between pathfinding and wayfinding now binds. She also declines to analyse the AI-lab CEOs when Lacqua raises Amodei, Altman and Musk: “I have not studied them specifically, but they don’t appear to be [traditional chief executives].”
Dynamic-capabilities tagging
digital-sensing/digital-mindset-crafting— the cell covers “establishing a long-term digital vision; enabling an entrepreneurial mindset.” Hill’s contribution is a boundary condition on the first half of that: for breakthrough change the long-term vision is not available (“many leaders have no vision about what’s really going to go on with AI”), and what substitutes is purpose plus wayfinding plus the declaration that “there is nothing called business as usual… everything is a working hypothesis.” The beginner’s-eye staffing rule — deliberately seating someone with no domain experience on the decision team — is the entrepreneurial-mindset half of the cell operationalised as a procedure rather than an exhortation.digital-seizing/strategic-agility— decision rights are the mechanism she names for “rapidly reallocating resources; accepting redirection and change.” The argument is explicitly about sensing latency (“somebody down there is going to hear it sooner than you”) resolved by pre-deciding who decides what, what the CEO wants to hear before a decision, and where the escalation guardrails sit.strategic-renewal/organizational-culture— creative abrasion is a culture-design claim: “you need to amplify difference, but many CEOs minimize difference”, with the leader’s job being to make conflict productive rather than to suppress it. The feedback-vacuum diagnosis and the Jobs-removes-himself anecdote are both about engineering the conditions under which people say what they think.contextual/internal-barriers— the cell names “high level of hierarchy; change resistances”, and Hill supplies a barrier the cell’s wording does not anticipate: the unspoken risk-appetite divergence at board level, sustained by conflict-avoidance and by the reputational cost of appearing “too conservative or resistant to the future.” This is a governance-layer barrier that blocks AI rollout decisions specifically, and it is the most concrete AI-related claim in the episode.
Linked entities and concepts
- Bloomberg Podcasts — publisher; second Leaders with Lacqua episode in the corpus. Updated in this ingest.
- Linda Hill — promoted to an entity page on this ingest, per the author-entity promotion rule: she was substantively cited but not an author in Ross & Schneider and flagged there for promotion on a second source. This is that second source, and here she is the principal speaker.
- durable-skills — creative abrasion and collaboration-as-the-decade’s-skill were previously held second-hand via Ross & Schneider; this is the primary statement.
- strategic-foresight — wayfinding vs. pathfinding as a stance for acting under irreducible uncertainty, distinct from scenario work that assumes a destination can be posited.
- dynamic-capabilities — the four cells tagged above, and the contextual-intelligence construct as a sensing microfoundation at the individual level.
- enterprise-ai-adoption — the unheld board risk-appetite conversation as a named blocker on AI rollout decisions.
- responsible-ai — risk appetite as something a board must state explicitly before deploying AI, and the ethical-decision reservation for humans.
- [[2026-05-24-erginbilgic-bloomberg-leaders-rolls-royce-turnaround-playbook|Erginbilgiç / Bloomberg Leaders]] — the practitioner counterpart from the same series; see
relationships:. - YC — the recorded disagreement about where irreducible judgement sits; see
relationships:. - Stanford GSB — the organization-theory grounding for the culture cell.
Dangling (single-source mention, deferred per author-entity promotion): Francine Lacqua (host; now on two Leaders episodes but as interviewer rather than author: — the wiki’s convention attributes video/podcast sources to the channel, so she is named in prose), Boris Groysberg, Nitin Nohria, Joe Mazzulla, Ed Catmull, John Kotter, Warren Bennis, Paradox Strategies, Harvard Business School, Cleveland Clinic Abu Dhabi, Pixar.
Source quality note
Human-curated captions; transcription is clean and only two proper nouns required correction. The cold-open duplication and the scripted voice-over segments are noted above so that timestamped quotations are not mis-attributed to the interview proper.
This is broadcast media, not research, and it is promotional in the ordinary way — a professor with a recent book and a consultancy, on a business-news podcast. There is no vendor relationship and no product being sold to the listener, so the vendor-source confidence cap does not apply; but the absence of method and evidence does. Every claim is an experienced researcher’s synthesis of her own fieldwork, delivered conversationally.
Its value to the wiki is a named vocabulary for the leadership problem the rest of the corpus keeps circling without naming. Across the enterprise-AI-return-gap cluster, the recurring finding is that leaders cannot say where AI is taking their organisation. Most sources treat that as a deficiency to be fixed with better strategy. Hill’s contribution is to say it is structurally correct for breakthrough innovation — that demanding a vision is the wrong request — and to name what replaces it: purpose rather than direction, working hypotheses rather than plans, deliberate beginner’s-eye staffing, pre-agreed decision rights, and an explicit board conversation about risk appetite that most boards avoid having. That reframing is worth holding even though nothing in the episode measures anything.