Making AI Work With You — Not For You
How the strongest operators stop using AI as a shortcut and start collaborating with it as a thought partner.
Most teams are using AI wrong — not because the prompts are bad, but because the relationship is. They're treating it like a faster typist. A summarizer. A task-doer. The teams getting outsized leverage are doing something different: they're using AI as a thinking partner.
I learned this in an environment where AI wasn't framed as a productivity tool. It was treated like a colleague. We didn't use it to do things — we used it to think better.
How My Perspective on AI Evolved
In a previous role at a high-growth company, AI was woven into how the team operated. We had a partnership with a leading AI lab — researchers sat in the office and ran weekly tutorials. We had a dedicated internal channel for sharing prompts, results, and emerging use cases across teams.
We weren't using AI to just do things. We used it to think better — to challenge our reasoning, surface blind spots, and pressure-test our work before anyone else saw it.
Use AI Like a Teammate — Not a Tool
The real value of AI shows up when you stop asking it to write faster and start asking it to think sharper. That mindset shift changes what you ask it to do — and more importantly, what you ask it to challenge.
"Where is my logic weak or underdeveloped?"
"Does this explanation make sense to someone outside my team?"
"How might someone misunderstand this?"
"Can you challenge my assumptions in this argument?"
"What are better ways to structure or pitch this?"
"If this were a presentation to leadership, what would they question?"
"The strongest strategists don't just prompt AI — they collaborate with it."
Reframing AI From a Shortcut to a Strategic Teammate
The difference between average and exceptional AI use isn't prompt engineering. It's posture. Most people use AI to compress time. The best operators use it to expand thinking.
The cycle isn't 'prompt and accept.' It's 'prompt, push back, refine, push back again.' Treat the output as a draft of your thinking — not a finished answer.
Why You Should Build an AI Assistant That Thinks Like You
Out-of-the-box AI tools are smart, but they don't know your voice, your judgment, or your way of working. That's why building a custom assistant — a GPT, a project workspace, a structured system prompt — isn't just helpful. It's the operational unlock.
Approaches problems through your lens, not a generic one.
Mirrors your tone, structure, and communication style.
Operates within your workflows, not around them.
Closes the gap between your expertise and execution — without losing the you in it.
What to Feed Your AI Assistant
To act like an actual assistant, it needs to know how you work. Treat onboarding your AI like onboarding a new hire — give it the context, artifacts, and operating rules it needs to think alongside you.
Formal JD plus any team playbooks that define your role and responsibilities.
Routine actions, workflows, and repeatable processes the assistant can support.
QBRs, strategy decks, and client-facing reports that show your standard of work.
Sent emails that train tone, structure, and client voice.
Who's on your team and what each person owns — so the assistant knows where work routes.
Client- or business-specific success metrics so recommendations are framed against what actually matters.
Always follow your organization's AI use policy. Avoid sharing confidential client data, sensitive personal information, or proprietary materials unless they've been explicitly cleared for AI use.
Use AI Strategically — Not Just Functionally
Most AI requests are functional: write this, summarize that, draft this slide. They're useful, but they're not the leverage. The leverage is the second prompt — the strategic one — that turns AI from an executor into an advisor.
"Write a follow-up email to the client based on these notes."
"What might the client push back on? What questions might this raise?"
"Summarize this transcript."
"Group these notes into: what the client wants, what we promised, and next steps by team."
"Draft a QBR outline based on these notes."
"What objections or alternative POVs should we address before presenting this?"
"Create a performance summary slide."
"What's the single takeaway we want leadership to act on — and what supporting evidence makes that case?"
"AI isn't replacing us. It's enabling us to do our job better."
The Operator's Takeaway
The shift isn't technical. It's cultural. The operators who win the next decade won't be the ones who prompt the fastest — they'll be the ones who built systems where AI thinks alongside them, inside their voice, inside their process, on the problems that actually matter.
Stop asking AI to do your work. Start asking it to challenge it.