You’ve heard of AI vibe coding, one dictionary’s phrase of the year for 2025. As of this week, 2026 is shaping up to be the year of the AI vibe shift.
You wouldn’t know the shift existed from the tech world’s top pronouncements of late; it is, after all, always sunny in Silicon Valley. Microsoft’s Build conference, like Google I/O in May, featured tons of techies talking about tokens, the metric by which AI prompts and answers are measured (a token, weirdly, is about three-quarters of a word on average).
Both conferences also centered claims about frontier AI that are dubious to say the least. DeepMind CEO Demis Hassabis at Google I/O: “Artificial General Intelligence is just a few years away… we are standing in the foothills of the Singularity.” Microsoft AI CEO Mustafa Suleyman: “scaling laws are holding… we are building towards what we call Humanist Superintelligence.”
Investors, too, showed little sign of losing their AI optimism this week. Nvidia stock tumbled for a few days, but rallied after CEO Jensen Huang insisted AI agents will run everything, everywhere in the future (presumably once they’ve stopped deleting databases). Anthropic, OpenAI, and SpaceX continue to chase trillion-dollar IPOs, the latter based in large part on the untested concept of AI data centers in space.
Support authors and subscribe to content
This is premium stuff. Subscribe to read the entire article.
