Episode 352

Google Next had a strange energy this year. Muted, almost subdued. Hard to tell if that’s general conference fatigue, world-events fatigue, or just a quiet anxiety that Google might announce something that makes half the room’s jobs irrelevant by lunch.
My bigger disappointment was that Google didn’t borrow a page from AWS’s Re:Invent playbook. That signature lightning round of broad, non-AI announcements gives the whole ecosystem something to chew on. Instead, Google buried the lead under a single thesis: everything is agentic. Full stop.
What was missing was depth. Real customer stories. Concrete use cases from people actually running these technologies in production, not polished slides about what’s coming. A lot of it felt uncomfortably close to vaporware, and that’s a problem when your competitors are shipping.
I think that’s exactly why OpenAI and Anthropic are out-innovating Google right now. They’re closer to their customers. Google has the infrastructure and the models, but the customer engagement gap is starting to show.
The other thread worth watching: Google is going heavy on Gemini Enterprise to go toe-to-toe with Copilot. Which raises the real strategic question. Do you win the AI war by embedding into existing productivity workflows where work already happens? Or is the future fully agentic and the smarter bet is investing in coding and tooling infrastructure to support that?
Both theses could be right. Time will tell.