Winning with AI
A Conversation with Charlene Li
My friend and longtime collaborator Charlene Li has just published a really important book called "Winning with AI: The 90-Day Blueprint for Success (co-authored with Dr. Katia Walsh), and I couldn’t resist asking her a few questions. I think you’ll find her answers as energizing as I did. This is a very important topic for our rapidly changing times!
SBK: Your book makes a counterintuitive claim: That organizations don’t need an AI strategy; they need AI to serve their existing strategy. But if AI is as transformative as everyone says, doesn’t it reshape strategy itself? How do you hold those two ideas together?
CL: It’s a real tension, and I hold it by thinking about what never changes. The mission of an organization — why it exists, who it serves, what value it creates — that doesn’t change because of AI. What AI changes is your ability to execute on that mission faster, at greater scale, with more precision than was ever possible before. So when I talk to CEOs, I’m not asking them to rethink their purpose. I’m asking them to look at their top strategic priorities and ask: where is AI the accelerant? That reframing alone changes everything about how they approach implementation.
SBK: I’m fascinated by your concept of the “superhuman” — people who combine AI’s cognitive infrastructure with what’s irreducibly human: empathy, intuition, judgment, wisdom. Maslow would have recognized this. But developing those capacities takes time and inner work. How do you reconcile that with your argument that speed is the new competitive moat?
CL: What I’ve seen in the research — and in practice with leaders across industries — is that speed and wisdom aren’t actually in conflict. They’re in conflict only when leaders treat AI as a decision-maker rather than a thinking partner. The leaders who move fastest with the best outcomes are the ones who use AI to compress information-gathering so that their human judgment — which is irreplaceable — can operate on better inputs, faster. Jeff Maggioncalda at Coursera called this “cognitive bootstrapping”: he’d form his own perspective first, then use AI to pressure-test it, surface blind spots, and accelerate the vetting process. The judgment is still entirely his. The speed comes from how he gets there.
SBK: The human capacity for self-reflection — noticing patterns in yourself, choosing to change based on what matters to you — is something you can’t automate. And yet the AI conversation often centers on productivity and efficiency. Is there a deeper argument in your book about what’s at stake for human identity when we hand over cognitive tasks?
CL: Yes, and it’s the part I care most deeply about. There’s a real risk that organizations use AI purely to drive efficiency — which is fine as far as it goes — and miss the more profound opportunity: AI offloading routine cognitive work could free humans to be more human. More empathetic. More reflective. More creative. More purposeful. I call the people who realize that potential “superhumans” — not because they’re superhuman in some sci-fi sense, but because they’re finally able to express the full depth of what makes us human, which we couldn’t access when rote tasks cognitively depleted us. But it doesn’t happen automatically. Leaders have to choose to design for it. That’s the leadership question that’s most underasked right now.
SBK: What’s the one question you wish every leader reading your book would sit with before they start implementing?
CL: “What are you trying to accomplish — and for whom?” Not “what can AI do?” Not “what are our competitors doing?” Just: what is this organization actually here to do, and who does winning serve? Leaders who start there end up making much better decisions about where to deploy AI and how to measure success. The ones who start with the technology almost always get lost.
Charlene Li and Dr. Katia Walsh’s Winning with AI: The 90-Day Blueprint for Success is available now at winningwithaibook.com.
Very interesting!
I've been using ChatGPT the past 6 months to streamline some of my solopreneur business processes and in the healthcare realm, with great success in each.
I'm firmly in charge, but to get ideas about how to more efficiently complete a task or have my clients complete tasks has been fun and incredibly helpful. I ask it nothing about my core work - that's all me and I don't need its help for that. More of my energy is available for my teaching and I've seen the impact.
With healthcare it's been wildly helpful. I've lived with AFib for over 12 years. I'm also neurodivergent - neurocomplex, to use the term I first heard from Lindsey Mackereth. Every time I've mentioned my highly sensitive nervous system as an important aspect of my being and approach to care, I've basically been dismissed by doctors.
5 months ago, I was facing daily AFib, suspected the medication at the time was worsening things, and only had one option left - a medication I was afraid of because it had toxic side effects.
The first time I used ChatGPT was for help in understanding a research article provided by my doctor on the medication. The result was life-altering.
My sensitive nervous system and focus on nervous system regulation were understood - finally. Filters through which to answer the questions the doctors wouldn't. No ego or power dynamics involved. Which meant highly customized information that helped me see that this med would likely be excellent for me. Just enough complexity to help my neurocomplex brain relax from the anxiety of uncertainty, but not so much as to be overwhelming. I took the med and it was like flipping a light switch - heart immediately calmed, it's been that way ever since, and I got my life back. Major big deal.
Plus, with this helpful tool, the pressure I always felt around trying to be understood and get info from doctors was relieved. I do still ultimately count on my doctor and am not replacing him - and shared how I was using ChatGPT. Success all around.
Using AI to navigate healthcare system challenges with billing and insurance is also incredibly helpful. I've been through loads of such challenges and AI helps by doing the cognitive and emotional translating.
I can "dump" a situation to it - express my frustration, share the details of what I need - and it helps strategize and drafts a professional sounding email or letter that gets the job done. I used to have to do all of that myself and it is a huge relief to get that help.
I still always bat around the language with it - "that sounds too flat and AI..." - the ultimate decision is clearly mine. But used this way I find it to be really helpful.
I agree with the points made by the author. Used thoughtfully and for the best purposes, I find AI to be quite helpful.
I understand a large part of how that speed is achieved. I had to apply the upgrade myself, while debating with AI last year. In effect, you gain processing time, in-the-moment, by removing all coping thoughts that usually need to be processed first.
A big part of enabling this loss of coping thoughts, is to adopt a balanced, but positive-first mindset. This means, you build inside yourself a low resistance path for new information to be processed within the moment, rather than having to leave some issues unresolved, which tends to create stress, and drags us out of the moment.
By increasing the available processing power, and by building an approach that counters challenges with positive-first analysis, a key self-transcendent state is achieved. This, I call, "Peak-Growth". This is where every moment is processed within that moment and true flow becomes a habit - One is able to give that moment one's maximum attention.
Also related is Constant Gratitude in the Moment, which leads to an organismic flow of peak-experiences that lead to further attention to the moment being needed, and is, I think, a pre-cursor to the above.