I did a very 2026 thing this week. I wrote an URfriendinHR AI agent.
Thank you for reading this post, don't forget to subscribe!Not a “replace humans” agent. More like a pop-up shop version of me that you can walk into, poke around, and leave with a slightly uncomfortable amount of self-awareness. She’s not fully deployed yet because, plot twist, keeping always-on agents running is not cheap. So right now she’s a limited-time exhibit. A 24-hour museum installation where the gift shop sells emotional intelligence and mild accountability.
And yes, I fed her my entire brand brain: 137 URfriendinHR articles as her knowledge base.
But the last time I flirted with the whole “digital me” concept, it was… a lot. Specifically, on October 14, I wrote about making a video clone and landing squarely in the Uncanny Valley. Watching “me-but-not-me” speak in my cadence was equal parts impressive and “why is my nervous system filing a complaint.” It was close enough to be convincing. Off enough to be eerie. So this time, I didn’t hand AI my face, I handed it my mascot.
Think: a stylized little mini-me with an oversized head and main-character energy, the kind you’d set on your desk to silently judge your meeting etiquette. It’s still me in spirit, but with a wink. Less “Is that… my soul?” and more “Okay fine, we’re doing a bit.”
And that SHIfT matters.
Because here’s what’s messing with my head (in the most professionally constructive way):
If AI can reflect our skills and knowledge back at us with frightening accuracy… what’s left that’s actually ours?
My current answer is still presence.
Not “I showed up to the meeting.” Presence like: judgment, restraint, context, the ability to tell the truth without lighting the room on fire. The human moment where you decide, “I could say the clever thing… but I’m going to say the useful thing.”
Also, personally?
This is exactly like texting a friend for validation and realizing midway through the conversation you’re really talking to yourself. Which is fine. Except now “yourself” can draft a performance improvement plan with a warm tone and bullet points.
So I’m curious (and I want the honest version, not the polite one):
Would you build an AI version of yourself for work? What would you let it do, and what is absolutely off-limits?
Drop it in the comments. I’m taking notes like it’s a requirements workshop.