Could AI have written this? Probably. But instead, it did a dozen other things that left me time to write it myself.
I know someone who assumes creative writing is AI-generated if the author isn't known to be particularly creative. It's a strange assumption to me. I see writing as an extension of articulation. If you can form a thought and communicate it, you have something to write about. The challenge is getting it out of your head and onto the page. That's where AI is useful.
I use it to organize messy thoughts, pull research together, explore alternative approaches, challenge my logic, and occasionally provide a very smart spell check.
Sometimes I need a sounding board. Sometimes I need someone to point out the obvious hole in my argument that I've somehow managed to miss. What I don't need is something that agrees with everything I say. Blind agreement isn't a thought partner. It's an echo chamber with excellent grammar.
But there's a distinction worth making here: using AI to help you think isn't the same as asking it to think for you. AI can make repetitive tasks faster, help identify patterns in information, and offer different ways to approach a problem. It can help communicate an idea more clearly or introduce a perspective we haven't considered. Those are useful contributions — not necessarily correct ones.
Sometimes an answer misses the mark because I haven't provided context that I have in my head. Sometimes I need to approach the question differently. And sometimes the answer is simply wrong. The tool can be useful without being right about everything. This becomes particularly important when we move from using AI to support our thinking to using it to produce the thinking itself.
AI can combine concepts, suggest connections, develop strategies, and generate something that looks entirely new. That can be useful, even surprising. But a novel combination of existing information isn't automatically an original insight, and a polished answer isn't necessarily a good one.
This is where experience matters. An experienced marketer can look at an AI-generated strategy and recognize what's missing, which assumptions need testing, and whether the recommendations make sense for the business. They can distinguish between something that sounds impressive and something that will actually work.
Someone without that experience might see the same polished output and assume the work is done. The difference isn't necessarily the tool. It's the ability to judge what the tool produces.
And that's why I don't think the answer is to stop using AI or to dismiss anything it touches as inauthentic. Sometimes a clearer email is all we need. Sometimes a fast first draft is exactly the right starting point. And sometimes the work requires more expertise, scrutiny, and original thinking than a tool can supply on its own. The point is knowing the difference.
Use AI to handle repetitive work. Use it to organize information, challenge your assumptions, and help you see what you might have missed. Ask it for alternatives. Tell it when you want pushback instead of agreement.
But bring your own experience, judgment, and point of view. Question the conclusions. Know what you believe and why. The goal isn't to prove that AI didn't touch your work. It's to understand what it contributed, what you contributed, and whether the result is actually worth the effort.
— AB