AI Enabled Me to do Two Weeks of Work in a Day. Here’s What Actually Happened.

Yesterday, I looked back at the end of the day and realized I had done roughly two weeks of work. Not metaphorically. Not in a motivational sense. Literally. In a single day, I:

• Wrote a full slide deck
• Built a working tool to estimate revenue growth for clients
• Drafted two product requirements documents for our platform
• Wrote a two-page marketing narrative
• Built slides for our fundraising deck
• Consumed hours of thought leadership content on the next phase of AI
• Ran six meetings as CEO of my company
• Reviewed strategy and product work with my team

And that’s just what I can remember. The interesting part wasn’t just the volume of activity. It was the feeling. I wasn’t exhausted, I was energized. But what I just described is the highlight reel, and there have been some important lessons learned along the way.

The Moment I Realized Something Had Changed

10 years ago, everything I just described would have required a team of analysts, a product manager, and a week of meetings. Now, it was just me and a conversation with AI.

But the conversation didn’t start clean. The first version of the revenue modeling tool came back structured in a way that felt generic. Sure, it was technically correct, but it didn’t fully capture the nuance of how we actually think about our clients’ businesses. I had to go back in three separate times before it evolved into something I could put into practice.

The prompt that finally worked wasn’t clever, it was specific and targeted. “Assume the client has inconsistent monthly revenue and model conservatively, flagging assumptions so I can challenge them.” One change in instruction changed the output entirely.

The slide deck went through five drafts before I got something I’d put in front of an investor. The first draft had the structure correct, but drafts two and three really refined the narrative arc.

Draft four was when I refined the framing of our competitive positioning, moving beyond more general language into something more specific to our actual differentiation.

Draft five was mine.

That’s the part that doesn’t make it into the productivity posts: iteration is the work. The AI doesn’t replace your judgment as much as it surfaces your judgment faster.

Where Human Input Elevated the Outcome

The marketing narrative was the hardest piece of the day. I came back to it the next morning and rewrote the opening from scratch: by hand, without the AI. The AI version was polished, and it had the right words in mostly the right order, but it needed a stronger point of view and more personalization to fully reflect my voice.

That’s a meaningful point of distinction.

The best version of the marketing narrative combined both. I wrote the opening paragraph myself. This was raw, unedited and pulled from my decades of experience. Then, I handed it off to AI and said “Keep this voice. Build out from here.” The result was something neither me nor AI could have produced alone.

What once required analysts, researchers, writers, designers, and developers can increasingly be orchestrated by a single curious person using AI as a thinking partner. Not replacing thinking. Amplifying it. AI isn’t just speeding up work. It is expanding how much one human mind can do in a day.

But here’s where executives need to be careful. Don’t use AI to stop thinking, use it to think more.

Curiosity Is the Real Superpower

One of the most powerful self-reinforcing systems we have as humans is our own curiosity. Don’t let that trait go dormant by becoming reliant on AI rather than working together with it to improve upon human foundations.

I see the way other executives are using AI. They ask a question, they get an answer, and they move on. The output is being accepted at face value, forwarded to a team, and finally inserted into a deck. The AI has then been reduced to a more basic autocomplete function.

This isn’t just an observation; it’s what researchers call agency decay: the gradual erosion of our capacity to form independent judgment when we consistently outsource our thinking. The brain, like any muscle, atrophies without resistance. And importantly, it can happen without you actively noticing.

That’s not how yesterday worked for me. Every answer I got prompted a new question, and that’s exactly how it should be every time. What if we modeled this differently? What assumption is this hiding? What would the skeptic in the room say about this? That level of curiosity still has to come from the human; it’s what guides how effectively you use the system. The AI just made it easier and faster to pursue.

How to Integrate Organic Thinking with AI Systems

The framework I’ve landed on didn’t come from a blog post. It came from a lot of draft reviews,
morning rewrites, and moments where I noticed the AI producing something that wasn’t quite fitting with my thoughts or beliefs.

Curiosity, practiced as a leadership discipline, is what makes the difference — not just asking more questions but staying genuinely open to what you don’t yet know, and willing to be proven wrong by the answer.

• Start with your own thinking. Before you prompt, write a sentence or two about what you actually believe. Write about your instinct, your angle and your concern. Use that as a foundation for your prompt. You’ll consistently get a fundamentally different output than if you start from a blank query.
• Treat the first response as a first draft, not a final answer. The value isn’t in what AI gives you. It’s in the gap between what it gives you and what you know should be there. That gap is where your judgment lives.
• Ask AI to challenge you, not just help you. Some of the most useful prompts I’ve used are: What’s the strongest argument against this? What am I not seeing? Flag the assumptions buried in this so I can stress test them. This approach only works if you’re genuinely willing to have your thinking challenged.
• Reserve something for yourself. The opening line, the core argument, the thing that only you could say; write that without help and then let the AI build around it.

The loop that worked for me yesterday wasn’t: prompt, accept, ship. It was: think, prompt. push back, refine, own. That extra friction in the middle is what this process is all about.

The Future Belongs to Builders

The same force that let me do two weeks of strategic work in a day is the force that lets a small business sell 24/7 without adding headcount. It’s not about efficiency. It’s about making things possible that weren’t possible before.

Yesterday I did two weeks of work in a day. Some of it needed further refinement. Some of it benefited from being revisited by a human the next morning. It didn’t feel like work, however, it felt like thinking at a higher level.

The most powerful people in organizations going forward may not be the ones with the

biggest teams. They may be the ones who know how to orchestrate intelligence: they think

clearly, ask great questions, iterate quickly, and collaborate with AI systems as partners.