Most People Use AI Wrong

The biggest problem with AI right now isn't that people don't use it. It's that the people who do use it are using it for the wrong things.

I talk to people about AI every day. The ones who tried it and gave up usually fall into two camps. Either they asked ChatGPT to write a poem for their mom's birthday and thought "that's it?" Or they saw someone on LinkedIn posting AI-generated thought leadership and decided the whole thing was a scam.

Both reactions are reasonable. Both are wrong.

The party trick problem

Most people's first experience with AI is a party trick. Someone shows them ChatGPT and says "ask it anything!" So they ask it something they already know the answer to, like who won the 1998 World Cup. The AI answers correctly. Cool. Then what?

This is like someone handing you a Swiss Army knife and you only ever use it as a mirror. Yes, the blade is shiny. That's not the point.

The party trick phase is where most people get stuck. They play with it for twenty minutes, generate a limerick about their dog, maybe ask it to explain quantum physics, and then never open it again. It felt fun but not useful. And fun-but-not-useful doesn't survive past the first week.

What AI is actually good at

Here's what I wish someone had told me when I first started using AI: it's not a search engine and it's not a content machine. It's a thinking partner that happens to have read everything.

The things AI is genuinely life-changing for are boring. Not viral-tweet boring. Actually boring. The kind of tasks you procrastinate on because they require mental energy but aren't intellectually stimulating.

Things like:

Drafting emails you've been avoiding. Not the easy ones. The hard ones. The email to your landlord about the broken heater. The follow-up to a job application. The message to a friend you've been out of touch with. You know what you want to say but the blank page stops you. Paste in some context, tell the AI the tone you want, and you'll have a draft in ten seconds. Edit it. Make it yours. Send it. That email you've been putting off for three weeks is done.

Making sense of complicated documents. Got a lease agreement? An insurance policy? A credit card statement with fees you don't understand? Paste it into ChatGPT and ask "explain this like I'm not a lawyer." I've saved myself hundreds of dollars by having AI flag things in contracts I would have missed.

Planning things that have lots of moving parts. Moving to a new apartment. Planning a trip with six people. Figuring out which health insurance plan actually makes sense for your situation. These aren't hard problems. They're annoying problems. They require tracking dozens of variables and comparing options. AI is perfect for this.

Learning stuff at your own pace. Forget about "AI tutors" as some future concept. Right now, today, you can paste a chapter from your textbook into Claude and say "quiz me on this." You can ask it to explain derivatives using only basketball analogies. You can have it generate practice problems at exactly your level. This works better than most paid tutoring.

None of this is sexy. None of it will get likes on Twitter. All of it will make your week measurably better.

The output trap

The second big mistake is treating AI like a content factory. People hear "AI can write for you" and immediately think about producing more stuff. More social media posts. More emails. More blog content. More, more, more.

This misses the point entirely. The goal isn't to produce more. The goal is to think better.

When I use AI to draft an email, I'm not outsourcing my communication. I'm getting past the blank page so I can focus on what I actually want to say. When I use it to break down a complex topic, I'm not replacing my understanding. I'm building a scaffold so I can learn faster.

The people who get the most value from AI aren't the ones generating the most output. They're the ones using it to make better decisions, faster. To see angles they missed. To stress-test their own thinking.

Try this: next time you have a big decision to make, explain your situation to ChatGPT and ask it to argue the other side. Ask it what you might be missing. Ask it to poke holes in your logic. This is worth more than a thousand generated social posts.

The one-prompt problem

Here's a pattern I see constantly. Someone opens ChatGPT, types one sentence, gets a mediocre response, and closes the tab. "AI isn't that smart."

That's like walking into a restaurant, mumbling "food" at the hostess, and being surprised when you don't get a great meal.

AI is a conversation. The first response is almost never the best one. The magic happens in the back-and-forth. "Make it shorter." "Actually, focus more on the cost angle." "What am I forgetting?" "Give me three alternatives." This is where AI starts being genuinely useful.

One prompt gets you a generic answer. Five prompts gets you something actually tailored to your situation. The difference is enormous.

Start with your actual problems

If you've tried AI and bounced off it, here's what I'd suggest. Don't open ChatGPT and stare at a blank box. Instead, wait until you hit a real problem in your real life. Something annoying. Something you're procrastinating on. Something confusing.

Then open it and describe your problem like you would to a smart friend. Not in robot-speak. Not in "prompt engineering" language. Just... talk to it.

"I need to write a complaint to my HOA about the parking situation. They keep giving me warnings even though I'm following the rules. Here's what happened..."

"My kid is struggling with fractions and I don't know how to explain it better than the textbook does. Can you give me a way to teach this using pizza slices or something?"

"I'm trying to decide between two job offers. Here are the details. What am I not thinking about?"

That's it. That's the whole technique. Bring a real problem, explain it like a human, and then have a conversation about it.

Why this matters

We're in a weird moment where the most useful technology in a generation is being dismissed by half the population because the other half won't stop hyping it for the wrong reasons.

AI isn't going to replace your job tomorrow. It's also not just a toy. It's a tool, and like any tool, it's only as useful as the problems you point it at.

The people who figure this out now are going to have a quiet, compounding advantage. Not because they're "prompt engineers" or AI experts. Because they stopped procrastinating on hard emails. Because they understood their insurance policy. Because they made better decisions with better information.

That's not flashy. But it's real.

If you want to talk about how AI could actually help with something specific in your life, reach out at jake@readlaboratories.com. No pitch, just a conversation.