Major Companies Reconsider AI Costs

Watch on YouTube ↗  |  May 31, 2026 at 16:09  |  10:48  |  Bloomberg Markets
Speakers
Gautam Mukunda — Research Fellow, Harvard Kennedy School; Lecturer, Yale School of Management

Summary

Gautam Mukunda discusses the potential for an AI bubble, noting that revolutionary technologies do not always generate profits, as seen with airlines and biotech. He highlights the huge capital expenditures and physical infrastructure challenges for AI, and warns that valuations like NVIDIA's may be overblown. He also touches on labor market fears and political pushback against data centers.

  • AI may be a revolutionary technology but could fail to make money for investors.
  • Comparisons to airlines and biotech as capital-intensive industries that destroyed shareholder value.
  • NVIDIA's market cap compared to all arable land in Australia as a sign of potential bubble.
  • Physical infrastructure (data centers, power, labor) is a major constraint for AI scaling.
  • Data centers face local political opposition despite economic benefits.
  • AI creates fear of job displacement, giving employers leverage.
  • The speaker suggests the AI boom could be a bubble while still having real technological impact.
Up Next