AI Isn’t Just Doing the Work. It’s Becoming You.
What version of you is your AI becoming?
I once asked my AI assistant (which I’d built to help me brainstorm, strategise, and think through decisions) what my biggest weaknesses were. Not as a test, but like you’d ask a friend who’s seen you on your best and worst days.
And it answered. Without hesitation.
“You rush the framing of your ideas.”
“You overexplain when trying to sound simple.”
“You call it momentum, but sometimes you’re just doing too much.”
It didn’t feel like criticism. It felt like clarity and like it had quietly been taking notes all along. And then that’s when it hit me:
It was learning me. My patterns. My defaults. My blind spots.
Here’s the paradox:
The more we automate, the more we need to ask who we’re automating. Because AI doesn’t just speed up tasks. It reflects and reinforces identity. It reinforces the version of you that you feed it.
What we thought were just workflows turned out to be mirrors, quietly reflecting versions of ourselves we’ve long since outgrown.
The Tools Don’t Forget, Even When You Do
Two weeks ago, someone booked a 1:1 with me out of frustration.
She’d built something powerful on our platform and wanted to know how to embed it into her site. The AI assistant she’d spoken to on our app told her we don’t support that.
But she’d seen me talk about it all over my socials. So she was confused as to why was the assistant shutting her down?
And that’s when it clicked: It wasn’t wrong. It was outdated and out of sync with who we are now. It had been trained on an earlier version of us, when embedding was still on the roadmap. When we hadn’t rolled out documentation. When we were still figuring it out.
How to Keep Your AI You
The more you grow, the more your AI needs to grow with you.
Here’s how I’m learning to do it (the hard way):
1. Audit the knowledge it’s pulling from
Start by asking: What is my assistant actually trained on? Old docs? Slack threads? Website copy from two versions ago? If your tone, values, or offer have evolved, you need to update the source, not just the outputs.
2. Document the traits you want to keep
List the three traits your AI should always reflect.
For me, it’s: direct, warm, and empowering. Then do a tone check. does your assistant still sound like that? Or like a corporate FAQ in disguise?
3. Inject new truth regularly
Every month (or sprint), review what’s changed including offers, voice, FAQs, priorities. If it’s important to you, it needs to be important to your AI.
4. Talk to your AI like a collaborator or teammate
Set house rules. Give it feedback. Ask it to justify its answers. The more context it gets from you, the better it will represent you.
5. Make it part of your identity practice
This isn’t just a technical task. It’s part brand, part therapy. Because when you update your AI, you’re not just aligning a tool but you’re clarifying who you’ve become.
You’re Evolving So Should Your Assistants and Agents.
We spend so much time trying to be clear in our messaging, our branding and our systems, but how often do we stop and check if we’re being clear to ourselves?
AI has a strange way of revealing that gap. It doesn’t argue. It doesn’t question. It simply repeats what you’ve told it whether you’ve done it explicitly or not.
So if it’s amplifying the version of you that second-guesses, overexplains, or stays small… It might be because that’s what it was shown.
The future isn’t just automated, it’s amplified. And the most powerful thing you can feed your AI isn’t data but rather, clarity.
Because when you build with intention, your assistant stops being just a tool. It becomes an extension of your growth. Of your values. Of your story that is still being written.
All the Zest 🍋
Cien
P.S. If this made you feel something, forward it to someone who might need a reminder: we’re still here, still dreaming, and now, maybe, a little closer to touching it.
Listen in: [Subscribe to the Podcast]