Working Better with LLMs: 10 Habits for Teams That Build

The biggest unlock is treating it less like a search box and more like a teammate.

I’ve been spending a lot of time thinking about how teams actually get better with LLMs. And honestly, I don’t think it’s about having the perfect prompt.

It’s more about building new habits.

The practical habits are simple:

Start your work with the frontier tools.
Give it a real brief.
Ask it to help you think, plan, review, test, and improve.
Use the right model for the job.
Let it take on the boring, repetitive stuff that drains time and energy.
Treat it less like a search box and more like a teammate.
Do not treat it as a replacement for your judgment.
Do not blindly trust the output.
Use the approved tools available to you.
Review the output and own the final work.

A hand-drawn top-ten infographic showing practical habits for teams working better with LLMs. — Better LLM work comes from team habits: brief clearly, use the right model, review, prove, and own the work.

That means treating it as a working partner that can help you move faster, see more options, and raise the quality of the final output.

I put together this top 10 list as a practical reminder for anyone trying to build better AI habits at work.

The point is not to wait until you’re an expert. Start with the work in front of you. Use the approved tools available to you. Keep learning by building.

And of course: don’t put confidential, restricted, personal, or regulated information into unapproved tools. Review the output. Own the final work.

Build.