Why the AI bubble isn’t a bubble (yet)

In this episode of 60/40, the crew goes into full AI group therapy mode.

Maria, Prashanth, Peter, and Alexey debate whether we’re living through another tech bubble—or just the warm-up act. From OpenAI’s trillion-dollar spend and AMD’s $90 billion chip deal, to SoftBank’s robotics binge and Deloitte’s AI “oops” moment, the team breaks down where the hype ends and the hard math begins.

They unpack why OpenAI might actually be too big to fail, how retail investors end up holding the bag, and why Asia’s energy constraints could make it the AI world’s blind spot. Along the way: humanoid robots, nuclear-powered data centers, and a few cuddly market bears.

Timestamps

00:00 – The Bearish Camp. Are we really in an AI bubble? Goldman data says not yet — AI capex still under 1% of global GDP.

02:30 – What If OpenAI Fails? A trillion-dollar spend, no profits, and systemic risk: is OpenAI too big to fail?

09:10 – Capacity vs Demand. AWS déjà vu — everyone’s building at once, but can demand keep up?

14:30 – Bubble Math. From 1890s railroads to 1990s fiber optics: why today’s AI hype is still early-stage.

18:10 – Round-Tripping & Retail Bags. Cross-investments, AMD warrants, and who really pays when the music stops.

25:20 – SoftBank’s Robot Empire. Masa bets $5.4B that humanoids will outlast the hype.

31:00 – AI in the Workplace. Deloitte’s $440K AI flop vs Bank of Singapore’s “10-day-to-1-hour” win.

40:50 – The Vibe Coding Hangover. Lovable’s crash, hype fatigue, and the next evolution of AI-assisted work.

42:55 – Powering the Future. Why Southeast Asia’s missing the data-center race — and that might be okay.

53:30 – Wrap-Up: Friendly Bears Edition. We’re still in 1999, not 2000. The last sucker hasn’t walked in yet.

Give us feedback or nominate topics & speakers at 6040podcast@techinasia.com.

60/40 is a bi-weekly podcast co-hosted by Prashanth Ranganathan, Peter Bithos, Aleksey Mironenko, and Maria Li.

A podcast by Tech in Asia, a member of The Business Times.