There's being overleveraged, and then there's whatever this is:

3d Edited

You're entitled to your own opinions, not your own maths. The hyperscalers have spent a trillion dollars to make $80 billion on AI. Anthropic and OpenAI are not worth $1 trillion each. Revenue growth of Amazon, Microsoft, Meta, Oracle, Anthropic and OpenAI would need to increase betwen 50,000 and 100,000x faster than current trajectories in order to just break even with data centre build out cost by 2030. For them to turn a *profit* on AI they would need to generate more than $2 trillion and as much as $5 trillion a year. These are already some of the most profitable firms in the world (not OpenAI and Anthropic, those are pure play grifters). You're telling me that Microsoft, Google, Amazon and Meta will manage - in three years - to essentially create additional versions of their most profitable tech businesses, from thin air and amid a saturated market? Do we think there are currently millions of businesses who never pay for software or SaaS or tech tools who are - again, within three years - suddenly going to realise they simply *must* pay hundreds of millions a year for AI? Where are the investors? This needs to be new money - where are all the funds who have sat on their hands throughout 18 years of post-GFC bull markets who will all simultaneously twig they need to invest hundreds of billions in AI? Meanwhile, while hyperscaler revenues are forecast in grow 15 percent annually this decade, capex is due to grow 20 percent a year. And with OpenAI and Anthropic (companies that this year will burn through $1.69 for every $1 they generate) set to go public, it is clear to me we are in for a financal collapse. As Joachim Klement puts it in the FT this week: "Viewed through this lens, the IPO of these AI companies is probably nothing more than a major transfer of investment risk from the current owners to retail investors, pension funds and others who are willing to buy the hype." (And, yes, this image does show Meta has spent $230 billion to make $3 billion on AI. Source: https://lnkd.in/ejvMAgUp) https://lnkd.in/eTuvJ6SD