In this video, I'm going to unpack what I see as the bad side of AWS's latest AI push at re:Invent—specifically around agents and mega‑scale infrastructure.

You'll hear a lot of polished narratives about "frontier agents," AgentCore, and AI Factories as if they're turnkey answers to every enterprise AI problem. I'm going to challenge that. To me, much of this feels like a rushed, "me too" sprint into the agentic hype cycle—an attempt to keep up with the fast followers rather than a confident, opinionated vision from the market leader. And when AWS looks reactive, it looks weaker than it actually is.

We'll talk about where branded agents like Kiro, the Security Agent, and the DevOps Agent risk becoming keynote theater instead of durable tools. We'll dig into why AgentCore may be more abstraction tax than simplification. And we'll look at AI Factories and ask whether most enterprises are really ready to run their own mini‑AWS in a basement.

If you're trying to separate signal from marketing noise, stay with me.