Three out of four tech job posts now ask for AI skills. Two years ago it was one in seven. Somebody RSVP'd on your behalf and forgot to mention it.

Web developer, and AI integrator. Product manager, but make it AI. Back-end engineers, data analysts, pretty much everyone at the table now has a plus-one they didn't choose: very confident, occasionally full of shit, and impossible to uninvite. Nobody handed you a seating chart, either. Do you sit them next to the customers? The CEO? What happens when they get a few drinks in and start giving the toast?

The loudest advice out there is to panic-enroll in the nearest ML course and learn to build a person from scratch. Cute. The real question is how much of this stranger you're actually responsible for.

I put it to Ethan Yang, a senior engineering manager who leads teams across web platforms, data engineering, and data infrastructure, which makes him the guy who has to deal with the plus-one when it misbehaves.

What we get into:

• The three things Ethan says make you "fluent" in AI, and why you can build them in about a month, no 4-year degree required

• What an AI hybrid role really looks like day to day for front-end devs, back-end engineers, PMs, and data analysts

• Why model APIs, prompt design, and RAG come first, and which cloud tools come second

• The ML rabbit holes you can skip unless you actually want to be an ML engineer

You were never asked to raise the plus-one. Just to seat them, watch them, and know when to call security.