AI can tell you what to eat. It can help you put together a workout plan. It can help you think through a business idea until you have a plan you genuinely believe could change your life.

Tomorrow you can wake up and do absolutely nothing with it.

That’s the part I keep coming back to. We’re getting access to incredibly useful tools, and it’s still possible to spend the whole day distracted, overwhelmed and jumping from one idea to the next.

I think improving yourself becomes more important as AI gets more powerful. Understanding what you want, what gets in your way and whether you actually follow through will determine a lot of what you get out of these tools.

A better answer doesn’t make the decision for you

You can have a sensible meal plan and still choose the cookies. You can understand the workout and still skip it. The advice can be good. You can agree with every word of it. There is still something you have to do.

That gap is familiar to me. For years, I had goals I believed in and plans I didn’t follow. I knew enough to make progress. I kept finding ways to postpone it.

Give that version of me a tool that can generate endless plans and I’m not convinced his life automatically improves. He might just have more things he’s excited about and hasn’t started.

More options give you more ways to avoid choosing

Every new possibility invites another question. Could I build this? Should I learn that? Is there a better way to do what I was doing yesterday?

Exploring can be useful. But if every interesting answer changes your direction, nothing gets enough time to develop.

I don’t want to become excellent at spotting opportunities and terrible at doing anything with them.

Being able to choose a direction and stay with it long enough to produce a result matters. So does knowing when you’re changing course for a good reason and when you’ve just reached the boring or uncomfortable part.

No prompt can settle that honestly if you won’t look at your own behavior.

Learn what gets in your way

This is why I care about self-accountability. I want to compare what I say matters to me with what I actually spend my time doing.

If I keep avoiding something, I want to understand why. Is the task unclear? Am I afraid of finding out I’m bad at it? Do I like the idea of the result more than I’m willing to do the work? Have I committed to more than I can fit into a day?

Those answers change what kind of help is useful. If the task is unclear, AI might help me break it down. If I’ve already broken it down five times, asking again is probably a way to avoid starting.

Knowing the difference is part of using the tool well.

Finish something before asking for more possibilities

Here’s the rule I’d start with: when you ask AI for help with a goal, choose one action from the answer and do it before asking for another plan.

Send the email. Try the first workout. Put a rough version in front of someone. Then come back with what happened.

Now you have something specific to work with. Someone didn’t understand the offer. The routine didn’t fit your schedule. The first attempt exposed a problem you hadn’t considered. Your next question can be based on experience.

I want to get better at that loop. Decide, act, look honestly at the result, adjust. I want to trust myself to use an opportunity when I see one.

AI gives me more ways to pursue the things I care about. I still have to decide what those things are and make time for them tomorrow.