When to Use AI and When Not To

I use AI every day. Literally every single day. I build with it, I write with it, I think with it. It's the most important tool in my business.

And I also think most people are using it wrong.

Not wrong in a "you're holding it wrong" way. Wrong in a "you've stopped trusting yourself" way.

The rule of thumb

Here it is. Simple. The one thing I wish every business owner would tape to their monitor:

If you're using AI to validate what you already know, you're using it right. If it's going against your gut and you're going with the AI anyway, you're using it wrong.

That's it. That's the whole framework.

Your gut feeling isn't irrational. It's the result of years of experience, pattern recognition, and context that no machine has access to. When you've been running a therapy practice for 15 years and the AI suggests something that doesn't feel right for your clients, your feeling is data. Real data. The AI doesn't know your clients. You do.

I say this as someone who builds AI tools for a living. Your instincts are not the thing AI replaces. Your instincts are the thing AI serves.

Amplify, don't replace

For content, AI should amplify your voice. Not make you more knowledgeable than you are. This is the line I draw with every client.

If you know your subject and AI helps you say it faster, clearer, more consistently, that's amplification. Good.

If you don't know your subject and AI generates authoritative-sounding content that makes you look like an expert you're not, that's replacement. Bad. And your audience will figure it out. Maybe not today. But the trust will erode because the foundation isn't real.

The same applies to everything: AI-generated strategy recommendations, AI-written emails, AI-produced reports. If there's a real person behind it with real knowledge and real opinions, AI is a multiplier. If the AI is doing the thinking for you, you've got a problem.

Where AI belongs

Let me be specific. Not philosophical. Specific.

Yes, use AI for:

• Drafting content you'll edit in your own voice
• Analyzing data to find patterns you'd miss manually
• Scheduling, booking, and administrative workflows
• Research compilation (gathering sources, not drawing conclusions)
• First-pass editing and proofreading
• Generating variations of something you've already created
• Building tools and automations for repetitive tasks

No, don't use AI for:

• Client conversations where trust and empathy matter
• Pricing decisions (that's gut, context, and relationship, not data)
• Therapy session notes (that's a human-trust task, full stop)
• Creative direction without human editing (AI doesn't have taste)
• Anything where being wrong has consequences you can't undo
• Replacing the thing that makes you, you

Maybe, with guardrails:

• Social media content. Only if you have a voice framework keeping it authentic
• Client communications. Only if a human reviews every message
• Business strategy. As a thinking partner, never as the decision-maker

The pattern: mundane tasks, yes. Human connection, no. The stuff in between needs a human in the loop.

The echo chamber problem

This one's important and nobody talks about it.

If your AI always agrees with you, you've configured it to be useless.

I see this constantly. People set up ChatGPT as a yes-machine. They ask leading questions. They reward agreement. They get frustrated when the AI pushes back. And then they end up with an expensive mirror that reflects their existing thinking back at them with better grammar.

That's not a tool. That's an echo chamber.

Configure your AI to challenge you. Ask it to find the weakness in your plan. Tell it to argue the other side. The whole point is to think better, not to feel validated. If you want validation, call a friend. If you want to think harder, use AI.

Opportunity AI vs. Efficiency AI

I draw a hard line between two kinds of AI use, and this distinction changes everything.

Efficiency AI does the same thing faster. Automates what you already do. Speeds up existing processes. Sounds great on paper. But here's the problem. Efficiency AI tends to replace people. If the only value of your AI implementation is "now we need fewer humans," you've built a cost-cutting tool, not a growth tool.

Opportunity AI creates new possibilities. Things you literally could not do before. A content system that lets a solo therapist publish consistent, voice-authentic content without becoming a writer. A client matching system that surfaces compatibility patterns a human couldn't track across 200 cases. An operational dashboard that gives a business owner real-time visibility into something they were previously guessing about.

Both have their place. But when I work with clients, I lead with Opportunity AI. Because that's where you get your time back instead of losing your team. That's where AI gives you capacity for more human connection, not less.

The real question

"Should my business use AI?" is the wrong question. Every business will use AI. That ship has sailed.

The real question is: where does AI make you more of what you already are, and where does it make you less?

Use it where it makes you more. Stop where it makes you less. And trust your gut on the difference. Because your gut is right more often than the AI is.