The Journal. Is the AI Boom… a Bubble?
302 snips
Oct 14, 2025 
Eliot Brown, a finance reporter specializing in the economics of the AI boom, joins to discuss whether the current surge in AI investment is sustainable or merely another tech bubble. He explores the staggering costs associated with building massive data centers, like the one in Texas, and examines the financial implications of AI revenue projections. Drawing parallels to historical infrastructure bubbles, he highlights the significant risks involved, and the pressure on companies to justify their hefty investments amidst uncertain returns.
AI Snips
Chapters
Transcript
Episode notes
On-Site View Of Abilene Data Center
- Berber Jin visited OpenAI's massive data-center construction near Abilene, Texas and described active construction and infrastructure work.
- The site will host supercomputers and, when finished, span the size of New York Central Park.
AI Data Centers Consume Huge Power
- Data-center capacity is measured in watts and OpenAI's Abilene site will need about 1.5 gigawatts of power.
- That level of demand approaches the output of major infrastructure like the Hoover Dam.
Gigawatt Costs Scale To Hundreds Of Billions
- OpenAI and Oracle announced expansion plans that could bring OpenAI to seven gigawatts of capacity.
- OpenAI estimates each gigawatt costs about $50 billion, implying multi-hundred-billion dollar buildouts.