4 medications. A printable comparison checklist. Here's how Helen sorted her Part D plans.

Today's Tip: Sorting Through Medicare Part D Plans

Open enrollment started on October 15. And I'd bet most of you are doing exactly what you did last year… staring at a list of plans, wondering how anyone is supposed to tell them apart.

Let me explain…

Helen Park is 69. She's a retired middle school English teacher in Columbus, Ohio. She takes four medications… a statin for cholesterol, lisinopril for blood pressure, metformin for type 2 diabetes, and omeprazole for acid reflux.

Last October, she spent an entire Saturday afternoon on Medicare.gov trying to compare Part D plans. Six browser tabs open. A legal pad full of scribbled numbers. And a headache by 3 p.m.

"I couldn't figure out which plans covered my generic metformin at the same copay," she told me. "They all looked identical until the fine print."

Sound familiar?

Here's what most people don't realize… the confusing part of Medicare open enrollment isn't the decision itself. It's the translation. You're reading plan summaries written in insurance jargon… tier structures, formulary lists, coverage gap rules, preferred pharmacy networks. Every plan presents the information a little differently.

And that's exactly where AI shines.

You don't need AI to pick your plan. That's your call. But you can use it to take all the confusing pieces… your medications, your pharmacy, your budget… and turn them into a clear, organized picture that finally makes sense.

Here's the prompt that does it…

When Helen tried this, she was surprised by what came back. The AI explained that her omeprazole… which she'd always assumed was a cheap generic… sometimes falls into a higher tier on certain plans because some formularies prefer a different acid reflux medication. That one detail could mean the difference between a $3 copay and a $40 copay. Every single month.

The AI also built her a checklist she could print. Five columns, four rows… one for each medication. She took that checklist to Medicare.gov, pulled up three plans in her zip code, and filled in the numbers in about 20 minutes.

No legal pad. No headache. No six browser tabs.

Now, I know what you're thinking… "Can I really trust AI with something this important?"

And that's the right question. Here's the honest answer: AI is not looking at live plan data. It doesn't know exactly which plans are available in your zip code or what each one charges for your specific drugs in 2027. That's what Medicare.gov's Plan Finder does. What AI does is help you understand the system… the tiers, the jargon, the hidden gotchas. It's like having a knowledgeable friend explain what you're looking at before you start shopping.

You wouldn't let a friend pick your plan for you. But you'd sure appreciate them sitting next to you and saying, "Here's what that line actually means."

Open enrollment runs through December 7th. You have time. But don't wait until December 6th.

Here's what to do today: open Gemini (or Claude, or Grok), paste in the prompt above with your medication list, and spend five minutes reading what comes back. Then bookmark Medicare.gov's Plan Finder for this weekend.

Five minutes of preparation now can save you hours of confusion… and potentially hundreds of dollars in 2027.

★ Quick Wins

Skill Builder: The Trust-But-Verify Method

Here's something that catches most people off guard… AI doesn't "know" things the way you and I do. It predicts the most likely next word based on patterns it learned during training. And sometimes, the most likely-sounding answer is completely wrong.

This is called a "hallucination." The AI isn't lying on purpose… it doesn't have intentions. But it can confidently tell you something that sounds perfectly reasonable and be dead wrong. It might cite a study that doesn't exist. Or quote a Medicare rule that was never written. Or give you a drug interaction it invented from scratch.

So how do you catch it? Watch for three red flags. First: extreme specificity out of nowhere. If AI suddenly gives you an exact percentage, date, or dollar figure that you didn't provide… ask where that number came from. Second: overly smooth confidence. If every answer sounds equally certain with no hedging and no "it depends"… be skeptical. Good answers often include caveats. Third: claims that are hard to Google. If AI says "according to a 2024 Johns Hopkins study," copy that exact phrase and search for it. If you can't find it in 30 seconds, it probably doesn't exist.

Here's a prompt you can save and use anytime you want to pressure-test an AI response. Paste it right after AI gives you an answer you're planning to act on…

Use this prompt every time AI gives you information you're going to act on… especially anything involving money, health, or legal decisions. Copy it. Save it in your phone's notes app. And paste it right after any response that makes you think, "Wait… is that right?"

The goal isn't to distrust AI. It's to trust it the right way… the same way you'd trust a sharp friend who occasionally mixes up the details. You listen. You appreciate the help. And then you double-check the important stuff before you act.