From Singularity To Tokenomics: The AI Narrative Just Hit A Serious Snag

The moment that should have been the news of last week, but wasn't, came on a podcast recording last Friday. Andrew Macdonald, Uber's president and chief operating officer, sat down for an interview and said the company is finding it "harder to justify" its AI spending. He couldn't draw a line, he said, between rising token consumption and useful product improvements. The link is not there yet. And he used a particular word for what was happening inside his own engineering organization: tokenmaxxing.

This wasn't from some random blogger, as awesome as some bloggers may be - it came from the number-two executive at a $200 billion company that has been publicly, aggressively bullish on AI deployment. And it arrived two weeks after Microsoft began winding down most of its internal Claude Code licenses because the token bills had become untenable – even for the company that put $13 billion into OpenAI and operates the Azure cloud that hosts most of Anthropic's compute.