HBR IdeaCast: Strategy Summit 2026 – Who’s Going to Succeed with AI?
Date: April 2, 2026
Host: Alison Beard (Harvard Business Review), Adi Ignatius (Editor-at-Large, HBR)
Guest: Andrew (Andy) McAfee, Principal Research Scientist, MIT Sloan & Co-founder, MIT Initiative on the Digital Economy
Episode Overview
This episode features a masterclass by Andy McAfee from the HBR Strategy Summit 2026, focusing on the critical question: "Who’s going to succeed with AI?" McAfee examines the massive uncertainty surrounding AI's economic impact, the rise of "geek organizations," and what separates winning companies from the rest in the AI era. The discussion moves through practical frameworks, organizational culture shifts, management changes, metrics, and talent strategies for AI adoption—with actionable guidance for leaders navigating the generative AI frontier.
Key Discussion Points & Insights
1. The Great AI Uncertainty
- Divergent expert opinions:
- Some, like Eric Brynjolfsson, argue that "the AI productivity takeoff is already here." Yet other voices (e.g., Daron Acemoglu) remain skeptical about measurable productivity gains.
- “We’re in this era of deep, deep uncertainty… Nobody knows anything definitive about where this is taking us.” – Andy McAfee (03:37)
- Three possible futures for AI (03:44):
- Ends scarcity; drives abundance (exponential growth)
- Dystopian collapse (“The Terminator” scenario)
- Incremental productivity increase (status quo)
- Implication:
- Organizations must navigate and make decisions amidst profound ambiguity.
2. Andy McAfee’s Three-Part Playbook for AI Success (06:00 – 13:00)
A. Commit & Make AI an Organizational Priority
- “You’ve got to make a bet... Commit as an organization to AI.” (06:33)
- Concrete suggestion: Set explicit OKRs for AI use, communicate expectations, and measure progress regularly.
B. Agile Over Waterfall – Learn by Doing
- The Waterfall approach (upfront detailed planning) is ill-suited for high-uncertainty environments.
- “The main problem with Waterfall is that it doesn’t work… you cannot plan your way to success when nobody knows anything.” (08:28)
- Agile/iterative cycles drive rapid learning and adaptability.
- Notable quote from Steve Jurvetson:
- “I can see dead companies. They're not responsive enough... We run circles around these planning-heavy incumbents.” (10:18)
- With modern AI tools, iteration cycles are shortening dramatically—from years to months or even weeks.
C. Spread Best Practices – Amplify Power Users
- AI adoption is uneven: heavy power users vs. curious or sidelined staff (11:53)
- Opportunity: Identify and broadcast these power users’ strategies/org hacks across the business to accelerate learning and reimagination.
3. Audience Q&A Highlights
Is AI Commitment an Article of Faith? (14:09)
Q (Adi Ignatius): Is betting on AI a leap of faith or techno-optimism?
A (Andy McAfee):
- Identifies AI as a ‘general purpose technology’—it improves fast, spawns complementary innovations, and diffuses broadly.
- “Yes, yes and yes. That’s where my… evidence falls.” (15:13)
AI as a Durable Source of Competitive Advantage (16:12 & 17:12)
Q: If AI is accessible to all, do competitive advantages erode?
A:
- Cost declines don’t eliminate performance gaps.
- “AI is absolutely not going to be the great competitive leveler. It’s going to make the distinctions between companies… much bigger than they are today.” (18:01)
- Top talent gets freed from “administrative busy work” to focus on innovation.
Impact of Software Agents on Work (19:43)
Q: What are “agents,” and how should we get comfortable managing them?
A:
- Agents are autonomous tech handling complex tasks with less supervision.
- “With appropriate guardrails... these technologies can now go do their thing for longer and longer periods... and give you a progress report.” (20:39)
- Rapid improvement is ongoing but requires oversight.
4. Measuring AI Success & Handling "Work Slop"
AI Metrics & KPIs for Knowledge Work (21:39)
- Track actual usage first: Is adoption broadening beyond power users?
- Tie usage to goal-oriented, outcome-focused KPIs, not just output volume.
- “For some categories, it’s straightforward to measure… for others, more subtle. But our ‘credibility revolution’ toolkit is very good at teasing this out.” (23:45)
“Work Slop” – AI-Generated Low-Value Content (24:46)
- Danger: AI churns out plausible but mediocre content, creating more work for others.
- Originates more in process-focused organizations (emphasize box-checking) than in outcome-focused cultures (emphasize results).
- “We don’t want slop wars going on inside the enterprise… In outcome-focused organizations, I expect that situation to be different. There your job is to get stuff done that makes a customer happy.” (26:15)
5. Building a More Agile Culture
Making the Shift from Waterfall to Agile (26:49)
- Understand Amazon’s “one-way vs. two-way doors:”
- Is a decision easily reversible (two-way) or not (one-way)?
- “For two-way doors, let people try stuff... Signal in every way that failure is OK as long as you learn something.” (27:48)
- Champions psychological safety for experimentation.
6. Talent, Hiring & Skills for the AI Era
Entry-Level Talent and Apprenticeship (28:42)
- Pulling back on junior hiring is a mistake:
- Loses the apprenticeship ladder (hands-on learning).
- Cuts off the supply of digital-native, AI-enthusiastic power users.
- “There’s a big demographic falloff… We tend to be more set in our ways and less willing to try crazy new things like AI [as we age].” (29:44)
- Organizations like IBM and Microsoft are ramping up entry-level hiring to keep the AI talent pipeline flowing.
Notable Quotes & Memorable Moments
- “Nobody knows anything. It’s too early in the rollout of modern AI… There’s reason to be optimistic about it, reason to be pessimistic.” – Andy McAfee quoting Derek Thompson (04:18)
- “Waterfall is a pledge on the part of everybody involved not to learn anything while doing the actual work.” – Citing Clay Shirky (08:37)
- “I sometimes think I have a sixth sense. I can see dead companies… They’re dead companies walking, but they don’t know it yet.” – Steve Jurvetson via Andy McAfee (10:15)
- "AI is not the great competitive leveler; it's the great gap-widener." (18:07)
- “If you cut off your entry level hiring, you are probably sacrificing future opportunities to learn and the skilled people of the future.” (29:50)
Timestamps for Important Segments
- 01:57 – Andy McAfee’s keynote: Framing the AI uncertainty
- 06:00 – The three-part AI playbook: Commit, Iterate, Diffuse
- 13:23 – Q&A: Is AI commitment faith or evidence?
- 16:12 – Competitive advantage in the AI era; implications for management roles
- 19:43 – Understanding software agents in practical terms
- 21:39 – Success metrics and KPIs for AI in knowledge work
- 24:46 – Work slop: The risk of low-value AI output and cultural solutions
- 26:49 – Practical steps for leaders to foster agility
- 28:42 – The importance of entry-level hiring and building the next generation of AI talent
Episode Takeaways
- Most companies are still discovering “how” to succeed with AI amidst high uncertainty; those that commit boldly, adopt agile mindsets, and democratize best practices will pull ahead.
- The strategic gap between organizations will widen, not shrink, as AI becomes more widespread.
- Leaders must foster cultures where experimentation (with feedback loops and psychological safety) is valued, junior talent is developed, and genuine impact (not busywork) is the yardstick for AI’s success.
For listeners and non-listeners alike, this episode delivers a robust framework for thriving with AI—centered on clarity, culture, and relentless learning.
