The $2 Trillion Question Is AI Worth It

21 Mar 2026 · 1 min

Watch on YouTube

0:52

In this monologue, the speaker critiques the massive capital outlays that large corporations are making on artificial intelligence, estimating that total global AI spending may exceed $2 trillion, with roughly $200 billion of incremental spend needed to achieve a reasonable return. He questions how AI systems will acquire genuine subject‑matter expertise beyond the capabilities of a mid‑level analyst, arguing that current AI primarily automates low‑skill tasks and that many of these capabilities are already available for free through services like Gemini and ChatGPT. Consequently, he doubts there is a viable path for corporations to generate sufficient revenue to justify the projected AI capex, suggesting that the incremental spend will not translate into proportional earnings.

Chapters

What data are they putting into this? How are they going to endow these machines with enough subject matter expertise to do something besides what a mid-level analyst can do? I think we're probably up over $2 trillion in spending, at least for these big companies, right? Like for the world, could be even more. So in aggregate, we're looking at $200 billion of incremental spending that's gotta go to AI to earn a decent return on that. And I think that right now there's no path to that. Don't get me wrong, AI is doing a great job of taking out lower-level jobs. Yes, we can get it to do those things, but that just means I'm not spending the money. And by the way, that was all free on Gemini and ChatGPT. Does it do enough to justify $200 billion minimum? I think an incremental spend, they're not gonna do much above that, which means you're not going to get the kind of revenue or spend to justify all the CapEx.