Bezos, Huang and Son Clash Over AI Costs, Jobs and Buildout

Jeff Bezos, Nvidia CEO Jensen Huang and SoftBank founder Masayoshi Son publicly debated AI costs, jobs and infrastructure as about $380 billion flowed into AI-related companies this year.

Roughly $140 billion came from investment-grade bond issuance, about $220 billion from venture capital and another $21 billion from high-yield credit, together accounting for roughly 64% of tracked capital flows. The financing has coincided with rising AI capital expenditures at major technology firms.

SoftBank pledged €75 billion ($87 billion) to develop roughly 5 gigawatts of AI data center capacity in France, an announcement made in Paris with France’s president.

At Nvidia’s GTC Taipei keynote, Jensen Huang rejected claims that AI is reducing employment, saying, “People are talking about AI decreasing jobs, it’s complete nonsense.” Masayoshi Son, speaking in Paris, estimated the current AI cycle could be “more than 10x, probably 50x bigger than dotcom.” Jeff Bezos called the surge an industrial bubble and said speculative excess can leave behind useful infrastructure when weaker projects fail.

An Amazon executive instructed staff to stop using AI for trivial tasks after the company reportedly consumed roughly $500 million worth of tokens in a single month. Other large technology firms have issued internal cautions about AI spending, and some hyperscalers report free cash flow near decade lows.

Matthew Sigel, a VanEck strategist, calculated that a flagship AI model can summarize a 500-page book for about $2.50 and questioned comparisons that place AI infrastructure costs at seven times those of legacy systems.

Analysts cited a forecast that hyperscalers would need about $7 trillion in additional revenue over the next three years to reach a 7% return on invested capital. Corporate earnings reports in the coming quarters will include details on AI-related spending and capital investment.