Using AI Without Letting It Think for You

This month's series has covered the risks of leaning on AI for reassurance and connection. Fairness requires the other half of the story: these tools are genuinely useful, and I use them myself. The question was never whether to use AI. It's where the line sits.

I'm Christopher Schuman, a board-certified psychiatric nurse practitioner serving patients in Texas and Washington. Here's my practical framework for using AI well, including how it shows up in health care.

Where AI Genuinely Shines

Think of AI as a very good language summary tool, an extremely efficient version of what searching the internet used to be. Used that way, it's excellent at day-to-day efficiency: finding the flight, comparing the options, digging up the answer that would have taken you twenty minutes of searching.

It's also legitimately good at overviews of unfamiliar territory. If you're not in medicine and you want to understand why cholesterol matters or what a medication actually does, AI can give you a solid orientation fast. That kind of health literacy, walking into an appointment with better questions, is a real benefit, and I'd rather have informed patients than intimidated ones.

And yes, health care providers use AI too. Used properly, it's a fast route to a clinical practice guideline or a national standard that would otherwise take real time to track down. Which brings up the rule that governs all of it.

Rule One: Check the Source

When information carries weight, when it's actionable, when getting it wrong has consequences, verify it. Ask where the information came from. Make sure it provides sources, then go read them yourself.

This is exactly how it works in medicine: if AI surfaces a guideline, the guideline gets read, not just the summary of it. The same discipline applies to legal questions, financial questions, and anything touching your health. AI can compress and oversummarize, its knowledge can be outdated, and it can state wrong things with complete confidence. Small stakes, like store hours? Mildly annoying when wrong. Big stakes? Check the source. Every time.

Rule Two: Don't Let It Think Critically for You

Here's the deeper line, and it's the one I'd put on a poster: let AI summarize for you, but don't let it think critically for you.

The best analogy is math. It's fine to let the calculator do multiplication the fast way, as long as you can still do it the long way. The calculator is then a speed tool, not a substitute for a capability you lost. The same holds for writing, analysis, and reasoning. Students who outsource every essay aren't saving time on a skill they have; they're skipping the struggle that builds the skill in the first place, and there's a purpose to some of that struggle.

A useful self-test: could you evaluate whether the AI's answer is wrong? If yes, you're using a tool. If no, and the stakes matter, you've outsourced your judgment to something with no accountability for the outcome.

Rule Three: Keep Human Decisions Human

The clearest boundary from everything this series has covered: decisions about your relationships, your health, and your life's direction need people who actually know you, your history, and your context. AI is not a philosopher, and it's not a judge that has weighed all the evidence, even when its confident tone suggests otherwise. Use it to prepare for the conversation. Don't let it replace the conversation.

The Practical Summary

Use AI freely for: summaries, overviews, drafts, efficiency, learning the landscape of something new, and preparing better questions for the professionals in your life. Use it carefully, with sources checked, for anything actionable about health, money, or law. Don't use it for: judgment calls about people, decisions with permanent consequences, or any thinking you couldn't check yourself.

The Bottom Line

AI rewards the person who wants the truth and flatters the person who wants agreement. Bring it good questions, verify what matters, and keep your own reasoning in the loop, and it's one of the best tools of the decade. Hand it the wheel, and you've traded your judgment for a confident voice that doesn't know you. The tool is fine. The line is yours to hold.

Want Help With the Questions AI Can't Answer?

New patient consultations available within 2-4 weeks.

Insurance: We accept most major insurance plans and offer transparent pricing for self-pay patients.

Crisis support: If you're experiencing a mental health emergency, please call 988 (Suicide & Crisis Lifeline) or go to your nearest emergency room.

Christopher A. Schuman, MSN, ARNP, PMHNP-BC, is a board-certified psychiatric nurse practitioner and founder of Veritas Behavioral Health, serving patients in Texas and Washington.