What's this about?
Wired argues that outside of engineers and AI early-adopters, ordinary people still aren't using AI "agents" — putting last week's viral "nobody actually uses agents" debate into the mainstream press.
What is an "AI agent," exactly?
An AI agent is AI that doesn't just answer, but *acts* — it takes a goal and carries out multi-step tasks on your behalf (booking, shopping, filing, coding) using tools. The gap the article names: builders are obsessed with agents, but normal people don't yet have a task where handing off control feels worth the trust and setup.
Why does it matter?
It reframes the frontier from "make models smarter" to "find the product/UX that makes agents worth using for regular people" — which several KOLs have called the most valuable open question of the year.