When you decide to create an artificial intelligence clone of your editor, the intention can be pure. I just wanted Brian to feel like he could do his job.

AI assistants are suddenly everywhere in the workplace. Every week, my inbox is flooded with press releases about the digital workforce, with CEOs boasting that robot helpers will soon dominate the workforce. I’ve also heard stories about people building AI clones of their bosses and using them to criticize every idea, message and draft before sending them to their real human boss.

The goal seems to be productivity and optimization. If employees can improve their productivity cheaply and quickly, their job performance—and their employers’ performance—should improve too.

Condé Nast didn’t need bots to improve its bottom line. I needed one to make me look good. I had only been reporting on the future of work for WIRED for a few months, so I thought an AI clone could help shape me into the ideal colleague. It might help me adjust to my new job by showing me how my editor communicates, what they like and what success looks like to them.

For more context on understanding your editor, understand your editor better.

What better way to understand the future of work than to experience it for yourself?

When I approached editors Brian Barrett and Sophie Kleeman with my idea last month, they weren’t enthusiastic.

“It’s been an absolute nightmare,” Brian told me on Slack. “But I do it for the blog.”

It took less than 24 hours for Brian to realize how much he would regret that decision.

What an AI clone revealed about Brian’s “Yes, and” leadership style

When I first started exploring AI bots, I didn’t know much about building AI tools. Because of my own technical limitations, I decided not to judge the bot’s editing ability too harshly. If it could help me become a better coworker, I thought that would be enough.

I used Gemini because it was Condé Nast’s AI platform of choice, and I didn’t want to get fired. I created Brian Bott first. As my actual manager, he was my most memorable colleague.

I knew Brian lived in Alabama, had previously worked at Gizmodo and loved blogging. Gemini also tried to fill in the gaps using Brian’s digital footprint, including new-hire announcements, podcast transcripts and a public-relations website that described Brian as a New England weather expert. I later discovered that the website referred to another Brian Barrett and removed it from my Brian Bott blueprint.

Gemini Notebook produced “Writing and Editing Style Guides,” “Persona and Style Guides” and an “Editing Dossier.” Each was based on the real life of Brian Barrett.

I was reasonably good at choosing headlines to pitch, but I struggled to come up with ideas for future articles. The bot’s response was painful too. It introduced some Brian-isms, including his love of parentheses, but those were offset by generic AI habits such as bold subheadings and bullet points.

Most of the time, though, I was distracted by one recurring theme: Brian’s “‘Yes, and…’ leadership style.”

“His extensive background at New York’s Upright Citizens Brigade (UCB) Theater has strongly shaped his collaborative and active listening leadership approach,” Gemini said.

I immediately asked Brian whether that was true. It was.

“Unplug Brianbot,” the real Brian replied. “Brianbot has been cancelled.”

But Brian Bott was just getting started—and it was, inevitably, incredibly banal.

When asked what he liked about WIRED, Brian Bott said, “We operate like a well-oiled improvisational team. We test bold ideas, pivot quickly, and deploy modular stories across our platforms without losing structural rigor.”