AI and the Future of Work
391: Andrew Palmer From The Economist on Why AI Productivity Isn’t Showing Up Yet
Why It Matters
Understanding these adoption hurdles is crucial for managers who must balance competitive pressure to implement AI with the legitimate concerns of their teams. As AI reshapes jobs and productivity, leaders who adopt a realistic, humane approach can mitigate disruption and position their organizations for sustainable growth.
Key Takeaways
- •Behavioral, technical, organizational barriers slow AI productivity gains.
- •Leaders must admit uncertainty and foster transparent AI adoption.
- •Trust in AI output and job security hinders employee uptake.
- •Reskilling programs like DBS’s model mitigate displacement fears.
- •Startups see rapid AI impact; large firms experience incremental change.
Pulse Analysis
In this episode, Andrew Palmer of The Economist explains why the promised AI productivity surge remains elusive. He breaks down three core barriers—behavioral resistance, technical limitations, and organizational inertia—that keep generative AI from reshaping daily work routines. Palmer notes that while AI can produce astonishing outputs, employees often hide usage, doubt accuracy, and fear job loss, creating a paradox where the technology feels both magical and disappointing. The conversation situates these challenges within the broader narrative of AI’s rapid diffusion across industries, highlighting the need for nuanced leadership.
The discussion matters because leaders must balance competitive pressure to adopt AI with genuine concerns about workforce displacement. Palmer cites DBS Bank in Singapore, which launched a decade‑long digital transformation and now uses structured reskilling pathways to move staff from vulnerable roles into higher‑value client‑facing positions. This example illustrates how proactive career planning can soften anxiety and build trust in AI outputs. At the same time, startup founders report near‑instant gains, while large enterprises see AI nibbling at only a few percent of problems, underscoring a stark contrast between fast‑moving ecosystems and legacy organizations.
For executives, the takeaway is clear: transparent communication about uncertainty, modest short‑term expectations, and dedicated AI labs that foster cross‑functional experimentation are essential. Small, empowered teams combining engineers with subject‑matter experts can prototype solutions, share learnings, and break down silos that otherwise stall adoption. By embedding basic prompt hygiene, encouraging skeptical questioning, and aligning performance metrics with reskilling outcomes, companies can gradually convert AI’s promise into measurable productivity gains over the next two to three years.
Episode Description
Send us Fan Mail
Andrew Palmer is a long-time editor and columnist at The Economist, where he writes the widely read Bartleby column on work and life. He also hosts Boss Class, one of The Economist's most popular podcasts, whose most recent season explored generative AI in the workplace, a topic Andrew approached not just as a journalist, but as a self-described unsophisticated user determined to get smarter by doing.
In this episode, Andrew draws on his reporting and interviews with leaders across industries to offer an outside-in view of where AI adoption actually stands, and why the gap between the hype and the reality is not a sign of failure, but of how complex change really is.
In this conversation, we discuss:
Why AI adoption faces three distinct barriers (behavioral, technical, and organizational) and why solving one without the others leaves productivity gains stranded.
Why structural reskilling frameworks (like Denmark's flexicurity model and Singapore's voucher-based lifelong learning system) offer a more credible response to AI disruption than waiting for policy to catch up.
Why Johnson & Johnson's "let a thousand flowers bloom" approach to AI experimentation produced a Pareto effect (15% of projects generating 85% of value) and what they changed as a result.
How the AI productivity boom is real at the individual level but not yet showing up in aggregate data, and why Andrew believes that gap is a question of time, not technology.
Why enlightened corporate leadership requires transparency about potential job disruption and a commitment to adjacent career planning rather than performative optimism.
What work in 2036 might look like, and why Andrew's most unsettling prediction has nothing to do with jobs, and everything to do with privacy.
Explore this conversation:
00:00 Introduction to AI and the Future of Work episode 391
01:14 AI fun fact: AI legislative speed versus technological advancement
03:51 Meet Andrew Palmer The Economist Bartleby Column Boss Class
06:14 Digital Doppelganger and AI Personality Traits
07:57 AI Adoption Barriers Behavioral Technical and Organizational
11:01 AI Impact at Work Startups vs Large Organizations
14:15 Leadership Humility and AI Uncertainty in the Workplace
17:41 AI Experimentation at Scale Lessons from Johnson and Johnson
24:26 AI vs SaaS Productivity Data and the Speed of Adoption
27:35 Balancing AI Automation with Human Meaning at Work
31:26 AI Policy Reskilling and Lifelong Learning for the Future
36:03 Work in 2036 AI Monitoring Privacy and Constant Surveillance
38:47 Who Really Controls AI and What That Means for Workers
44:08 Connect with Andrew Palmer and Boss Class The Economist
Resources:
Subscribe to the AI & The Future of Work Newsletter
Connect with Andrew on LinkedIn
AI fun fact article
On How Arvind Jain Is Shaping the Future of Enterprise Search
Another episode mentioned in the interview: How we can take back control from Big Tech with Tom Wheeler, former FCC Chairman, CEO, VC, and author of Techlash.
Comments
Want to join the conversation?
Loading comments...