I find myself explaining this almost every week. In client meetings, on advisory calls, in conversations with other founders. The same questions, the same fears, the same misunderstandings about what AI actually is and what it's good for.
So here's what I keep coming back to, after two-plus years of working with these tools daily, alongside a team that's pushing them into real work across strategy, creative, and technology.
AI is a tool. You are the expert.
That's the order of operations. If you flip it, you're in trouble.
If you know nothing about a subject, AI can be helpful or it can be completely dangerous. It sounds confident whether it's right or wrong. It doesn't pause. It doesn't second-guess. It just keeps going. But if you have real experience in a space, AI can be extraordinary. It becomes a research partner that never gets tired. You ask the question, it works the problem. You ask a better question, it goes deeper.
The difference between good and bad AI usage is not the tool. It's the person holding it.
Think about it like a creative review. Three designers hand me comps for a client project. One is completely off the mark. How do I know? Because I know the client. I know what we sold. I know what we were trying to produce. I have context the work gets measured against. AI is the same way. You give it information and it comes back with its own take. Sometimes it's sharp. Sometimes it's way off. Sometimes it surfaces something you never considered and you go do your own research and come back thinking differently. Every good interaction with AI should leave you a little smarter than you were before.
But if you don't have experience, you can't spot the difference between the good output and the garbage. You just take it at face value. That's where it gets dangerous.
AI gathers and organizes. You decide what's real. You decide what's meaningful. You create the work that actually moves things forward. That line isn't going to blur anytime soon.
Elevation, not replacement
I shared something with my team recently that I want to share more broadly. Your role is not at risk because of AI. It is elevated by it.
That's not a corporate talking point. I believe it because I've watched it happen. We ran an internal exercise where the team had to use AI to solve a creative problem. The person who cracked it wasn't the most senior person in the room. It was someone doing ground-level execution work. Nobody else even got close.
Title doesn't predict who will thrive with this. Mindset does.
I've heard from people across industries that some of their colleagues won't touch AI. They're afraid that learning it signals their role can be automated. I understand the instinct. But it's backwards. The people who refuse to engage are the ones most likely to be displaced. Not because AI replaced them, but because they stood still while everyone around them adapted.
People compare AI to an intern. It can function at that level. You ask it to bring you a cup of tea. It comes back with half a cup. You direct it. You correct it. You refine. But AI doesn't stay at the intern level unless you do. When you bring real experience to the table and learn to direct with precision, it goes well beyond fetching tea. It surfaces patterns, references, and angles you didn't have time to research on your own.
The ceiling isn't the tool. The ceiling is you.
The industrial revolution, not the apocalypse
Every week someone asks me if I'm worried about AI taking jobs. Not entirely. But I'm not dismissive about it either.
AI is the industrial revolution of our time. Processes will change. Entire workflows will shift. Some roles will look completely different in five years. Some people will be displaced, at least temporarily, while things realign. That's real, and pretending otherwise doesn't help anyone.
But people will work. People will figure it out. New types of work will emerge that we can't fully see yet. The same fear existed during every major shift in how work gets done. And what actually happened every single time was adaptation, not elimination.
The change is real. The panic is disproportionate.
Look at the news right now. It's designed to make you feel desperate. AI is going to take your job. The economy is collapsing. Everything is on fire. That's not a strategy. That's noise. Meanwhile, I recently got off a call with a colleague who has bombs landing three kilometers from his house. Real bombs. He's worried about his family's safety, not whether a chatbot is coming for his career. We have the luxury of worrying about AI. Let's use that privilege to think clearly instead of reacting from fear.
Governance and data security are real, especially in regulated industries. If you're handling client data, you need controls. You need to know where that data goes and who learns from it. But governance can't become a reason to do nothing. Use general AI tools with anonymized data to expand your thinking. When it comes time for deep analytics against proprietary client data, that's where private tools and strict protocols make sense. It doesn't have to be all or nothing. The people who treat it that way are going to fall behind the people who learn to work with both realities at the same time.
You stay in control
Here's the thing I come back to every single day. Nothing leaves without a human reviewing it.
AI can suggest. It can draft. It can compile. It can research. It can organize. It can surface things you never would have found on your own. But you accept, reject, or modify. That is the process. Every single time.
That's not a limitation. That's the whole point.
The most important thing in any room where AI is being discussed is not the technology. It's the people. Your expertise becomes more powerful with these tools, not less. Your judgment matters more, not less. Your ability to think critically, to challenge what's in front of you, to know the difference between insight and noise, that's the skill that makes everything else work.
So stop worrying about whether AI is going to replace you. Start asking how you're going to use it to become someone who can't be replaced.
That's the only question that actually matters.