Key insights
- The author argues that Google's massive $80 billion share issuance for AI investment, followed by Berkshire Hathaway's $10 billion purchase, signals an extremely expensive AI race. This mega-capex increase, potentially without clear ROI, suggests valuations for AI-related companies ('picks and shovels') are overblown, indicating a potential bubble nearing its peak and a subsequent market downturn.
As you all know, Google recently made decision to issue shares in ballpark of 80b usd to raise capital for AI investment. It’s so huge that they didn’t sell existing shares, they issued new. They are starved. Berkshire followed thru with this idea and bought 10b. They did time high price of stock to do this which goes in their favor for this move.
Now my only assumption is that they convinced Berkshire that they have clear path to ROI or more likely that this capex is mandatory to stay on top of things even if return is minimal, if any.
Whatever the case is, what is clear now is that AI game is extraordinarily expensive and that capex is near the top unless there’s serious ROI going to start happening right now, which I doubt.
This general mega capex increase can’t sustain without canibalising the companies. Therefore valuations of picks and shovels companies are overblown. Basically, we’re in a bubble that’s about to burst.