
Sign up to save your podcasts
Or


Picture a general counsel with a spreadsheet: columns for federal, all 50 states, the EU AI Act, and every jurisdiction her company touches. She’s trying to build a governance program on the one thing guaranteed to move, the rulebook. She’s going to lose, and not because she isn’t smart. In this episode, I explain why chasing AI regulation is a losing game and what to do instead: set internal standards that sit above any rule, using the AI Trust Pyramid from Winning with AI. I also share what I saw as a member of the Singapore Ministry of Health’s regulatory advisory board, where the regulator itself is taking a principles-first approach. Governance isn’t a speed bump. It’s the brakes that let you go fast.
I didn't plan to talk about sustainability at a recent keynote I gave for the American Institute of Architects. Turned out it was the only thing anyone wanted to talk about afterward. So today I'm tackling the question head-on: is AI bad for the planet? My answer isn't "stop using AI" — it's "use it lighter." I walk through the real numbers on data center water and power use, then break down four practical moves any leader can make: right-sizing your models (you don't need a frontier model to draft an email), using small and edge models that run locally, building context graphs instead of re-feeding entire documents, and fixing "prompt thrashing" through better batching. The throughline: sustainability isn't separate from responsible AI — it is responsible AI.
Everyone's excited about AI. Everyone has a ChatGPT license. And nothing is actually changing. Charlene calls this the AI hesitancy gap — the distance between what leaders know AI can do and what they're actually confident enough to decide. In this episode, she breaks down why "knowing, doing, leading" has the sequence backwards (it should be knowing, leading, doing), why adoption isn't the same as adaptation, and why 95% of AI pilots fail — not because the tech doesn't work, but because leadership abandoned them. Plus: the playground paradox, why pre-mortems beat pep talks, and how Goldilocks Governance turns judgment into an organizational muscle instead of a personal risk.
Everyone in the C-suite thinks they should own AI — the CIO, COO, CFO, chief risk officer, CHRO, and CDO can all make a compelling case. Charlene explains why that's exactly the problem, and why chief AI officers have been failing across organizations for two years. Drawing on the Four Building Blocks framework from Winning with AI, she introduces the "conductor" model: one person whose sole job is to coordinate AI's outcomes, roadmap ("the score"), and pace ("the tempo") across every department — without owning any single piece of it. Using an orchestra metaphor and the real-world example of Sandy Carter's rise to Chief Business Officer at Unstoppable Domains, Charlene lays out what the conductor role actually owns, when a CEO can (and can't) play it, and the three questions every board should be asking about AI leadership right now.
Disruption isn't coming, it's here.
Every week, strategist and author Charlene Li gives leaders the frameworks, the honesty, and the real-world perspective to lead through AI, organizational change, and conditions no one trained them for.
Short episodes. No filler.
If your palms are a little sweaty, good.
From the publisher's feed