Good Results, Humans Required?

Jan 28, 2026
Katelyn Reilly
Newsletter

Hello from Steyer,

Unfortunately, I do love reality tv. I’ve watched what I’d say is “far too much” of it in my life, if I were feeling particularly judgmental. I confess I love the relationships, the psychological landscapes making their way through the competition. For certain shows (looking at you, Love Island franchise), I love the self-aware, witty narration, the very existence of “the villain edit,” and the post-season social media parsing out of what, exactly, was tv versus reality. I’m also keenly interested in generative AI. So when I heard about a fully AI-generated reality tv survival show on YouTube (thank you for the link, Tess!) I was all in. I’m three episodes deep, and whew: it’s terrible. It’s great. It’s… a whole lot to think about.

Screenshot 2026-01-28 at 2.07.20 PM

Tess’ big question, when she brought this to our Tuesday AI Chat this week, was about the editorial decision-making involved in creating the show. When the credits roll, it’s a list of prompts the AI studio used. The AI-generated characters in the show (as explained by the AI-generated host) decide their courses of action for themselves, but just given the watchability of the show, it’s clear a human with visual storytelling experience and creative discernment has been in the mix. The “camera” tracks someone at a certain angle; the framing and pacing and constraints are just so. I want to meet this person, as I have “aggressive polar bear”-related lines of inquiry to pursue—and because the human vision behind the studio’s experiment is what makes it worth my time.

Related (I promise), there’s a good LinkedIn essay going around in which a vice president at a NZ law firm explains why his firm’s use of AI doesn’t result in steep client discounts. I’ve written here previously about the death of the billable hour, but exactly how to articulate and then calculate value in a world augmented with AI is far from clear-cut at this moment in time. Richard Burcher’s take on value, spoken to his clients: “You are not buying six-minute units. You are buying outcomes, resolution, clarity, risk mitigation, commercial opportunity, and — perhaps most importantly — peace of mind. You are buying the transfer of risk and responsibility from your organization to ours. That responsibility does not get cheaper because the tools improved.”

All of this to say: valuable results in the real world require discernment and trust. So that’s what Steyer will be doubling down on this year: maintaining our high quality bar for service, earning client and coworkers’ trust through integrity and accountability. (And I really hope we can achieve these aims in a way that won’t get us eaten by an AI-generated polar bear.)

Thanks for being here,
Katelyn