The Light Bulb Moment
I'm sitting at my friend Rolf's house right now. We're about to head to dinner — sharing a birthday celebration, because his was recent and mine is coming up, and at a certain point in life you stop doing separate parties and just combine them. It's more efficient and frankly the wine is better when someone else is paying.
Rolf has been in the yacht business for thirty years. High-end marine, rich-guy-toy business. Boats, deals, clients who buy things that cost more than houses. He knows that world cold. He understands markets, customers, money, logistics, and the human psychology of selling expensive things to people who want expensive things.
He does not understand the technology the way I do. And that's fine — that's actually the point.
We spent the afternoon talking about AI. Not in the abstract, Twitter-thread, "the future of work" sense. In the concrete, "what can this thing actually do for someone like you" sense. And I watched him go through something I've seen before but that never gets old to watch. It's the light bulb moment.
The light bulb moment is not when someone learns what a large language model is. It's not when they see a cool demo. It's not when they're impressed by a chatbot answering a question.
The light bulb moment is when someone who is genuinely smart — someone who has built businesses, managed people, closed deals, navigated complex markets — suddenly realizes that the bottleneck was never the tool.
The bottleneck was the clarity of their own intent.
Here's what I mean. Rolf has spent thirty years developing something valuable: the ability to look at a situation, understand what actually matters, figure out what needs to happen, and communicate that to people who can execute. That's what a good operator does. That's what a good CEO does. That's what a good dealmaker does. The skill is not in the hands. The skill is in the head. It's in the ability to structure a problem, see the path through it, and articulate the instruction that makes someone else able to act on it.
For thirty years, the people receiving those instructions were other humans. Sales teams, operations people, logistics coordinators, designers, engineers, accountants. You describe what you want, they go do it, you review the work, you adjust, you iterate. That cycle has a speed limit, and the speed limit is human. What I watched click for Rolf this afternoon is that the cycle doesn't have that speed limit anymore. Or more precisely: the cycle still has a speed limit, but it's no longer the execution side. The speed limit is now how fast you can think.
The thing I keep trying to explain to people — and that I watched Rolf actually get today — is that AI is not a tool you learn to use. It's not software you master. It's not a platform you adopt. It's an intelligence you instruct. And the quality of what it produces is almost entirely a function of how precisely you can describe what you want. Not in a technical sense. In a strategic sense. In a "I know exactly what outcome I'm looking for and I can articulate the structure of that outcome clearly enough that a machine can run with it" sense.
That is a skill Rolf already has. He's been developing it for thirty years. He just didn't know that's what it was, because for thirty years the other half of the equation — the execution — required a team of humans. Now it doesn't. Or rather, now it requires fewer humans, working alongside systems that can take a clear instruction and run it at a speed and scale that no team of humans can match. The light bulb moment is when you realize the thing you've been practicing for thirty years — structuring intent, communicating it clearly, reviewing the output, iterating — is the exact skill that matters now. The execution layer has been automated. The thinking layer hasn't. And the thinking layer is where you were already good.
I think this is why a lot of very capable people are confused right now. They're approaching AI the way they approached previous technology waves: what software do I need to learn? What platform do I need to adopt? What tool do I need to buy? And they're frustrated because none of those framings quite fit. AI doesn't behave like software you learn. It behaves like a colleague you direct. A very fast, very capable, occasionally weird colleague who never sleeps and will do exactly what you tell it to, including when what you told it was wrong. The people who get it first are not the most technical people. They're the people who are best at thinking clearly about what they actually want. The operators. The strategists. The people who have spent years developing the skill of looking at a messy situation and figuring out the through-line.
That's Rolf. Thirty years of yacht deals, client relationships, and high-stakes negotiations gave him exactly the skill that matters now. He just hadn't framed it that way, because for thirty years there was no machine on the other end of the instruction. Now there is.
So here's what I told him over the course of the afternoon, and what I'll tell you too:
Stop thinking about AI as something you need to learn. Start thinking about it as something you need to direct. The quality of your output is going to be determined by the quality of your thinking, not the sophistication of your tools. A person who can structure a clear, complete, well-reasoned instruction is going to get extraordinary results from AI. A person who can't is going to get garbage and blame the model.
The model is not the bottleneck. You are. But in a good way. Because the thing that makes you the bottleneck — the clarity of your own intent — is the thing you can actually improve. And you've probably been improving it your whole career without knowing that's what you were doing. Happy birthday, Rolf. The light bulb is on. Let's go eat.
Rich Washburn is a technologist and strategist working at the intersection of AI, infrastructure, and capital. He is Managing Partner and Chief AI Officer at Eliakim Capital and currently at Rolf's house, pouring another glass and getting ready for dinner.
Comments