I have written before about how AI has not replaced my CPA — it has made me a far better client. This year I got to test that theory under real pressure. And it held up better than I expected.
Here is what happened. Tax season arrived, my preparers were doing their work, and the clock kept ticking toward the deadline. My 2025 estimated results were not going to land in time. I had to file an extension.
If you have never been in that spot, here is the uncomfortable part: an extension gives you more time to *file*, but it does not give you more time to *pay*. You are still expected to estimate what you owe and put money down by the deadline. So I was sitting there with a real obligation and no number. That is a bad combination — you are either guessing high and parking cash you do not need to part with, or guessing low and setting yourself up for penalties and interest later.
I did not want to fly blind. So I went deeper than I ever have before.
The Pattern, Again
I took every tax document I had — the whole pile of PDFs, the forms, the brokerage statements, all of it — and I uploaded the lot to Claude. Then I asked it to help me estimate my 2025 taxes myself.
I want to be precise about what this was and what it was not. I was not asking the AI to *be* my accountant. I was asking it to help me read my own situation — to organize the inputs, walk me through how the pieces fit together, and produce a defensible estimate I could actually stand behind while I waited on the professionals.
We worked through it the way I would work through any hard problem: methodically. What are the income sources. How does each one get treated. Where are the moving parts. What do I genuinely not understand here, and can you explain it in plain English before we move on. By the end I did not just have a number — I understood *why* it was the number.
The point was never to skip the professional. It was to stop showing up empty-handed. By the time my preparers finished, I was not a passive recipient of a result. I was someone who could check their work because I had already done the homework.
The Moment of Truth
Then the preparers delivered their final number.
Claude's estimate was super close. Close enough that the gap did not matter for any decision I needed to make. Close enough that, during the weeks I was waiting, I had been able to plan and pay with real confidence instead of crossing my fingers.
That is the whole game, by the way. Not "the AI got the exact right answer to the penny." Tax is full of judgment calls and edge cases, and that is exactly why I keep professionals in the loop. The win is that I closed the gap between *no information* and *good-enough information* to make sound decisions in real time. The extension stopped being a source of anxiety and became a non-event.
I have come to think of this as the difference between flying blind and flying with instruments. The instruments are not the pilot. But you would never want to fly without them.
Then I Looked Forward
Here is where it actually started paying for itself.
Once I trusted the picture for 2025, I turned to 2026. Specifically, my estimated quarterly payments — the payments a lot of us with non-simple income have to make four times a year, where you are essentially pre-paying your tax bill in installments.
Most people, myself included in past years, treat these as a thing you sort of guess at and hope works out. You either copy last year's numbers or you follow whatever your software spits out, and you do not think hard about it. That is a mistake, because there is real money sitting in how you size and time those payments.
So I had Claude help me model it properly. We looked at the actual mechanics — how the quarterly system works, what the "safe harbor" rules are (the thresholds that, if you hit them, generally protect you from underpayment penalties even if your final bill comes in higher), and how to right-size each payment instead of blindly overpaying to be "safe."
Through smarter modeling — paying the appropriate amount instead of an inflated just-in-case amount, and understanding the rules well enough to be confident in it — I am on track to save thousands of dollars this year. Not through any clever trick. Just through actually understanding the system I was operating in.
That is the part I keep coming back to. The savings did not come from a loophole. They came from finally doing the homework on something I had been winging for years.
A necessary aside, because I take this seriously: none of this is tax advice, and I am not the person to take it from. The safe-harbor rules, what counts as income, how your specific situation works — that is exactly the territory where you want a real professional who knows your full picture. What I am describing is how I got *prepared* for those conversations, not how I replaced them. If your situation is anything like complex, talk to your own CPA. I did.
Why This Keeps Working
The reason this approach holds up across taxes, financial planning, even health stuff, is that the bottleneck was never access to expertise. The bottleneck was the homework gap — the hours of preparation it used to take to walk into a professional's office actually understanding your own situation.
Most of us never closed that gap. We did not have the time, and frankly the material was intimidating enough that we did not want to. So we showed up with a shoebox of PDFs, nodded at recommendations we half-understood, and outsourced not just the work but the *understanding* of the work. That is a fragile way to manage the most important things in your life.
AI does not close that gap all the way. But it closes enough of it to flip the dynamic completely. I now walk in knowing what I am looking at, what the stakes are, and what questions are worth asking. My professionals are better leveraged because they are not spending their expensive hours getting me up to speed — they are spending them on the judgment calls that actually require a human expert.
The instruments are not the pilot. But this year, for the first time, I had instruments. And the difference was not subtle.
Try It On Your Own Stack
If you have been treating your taxes — or your finances, or anything else you pay a professional to handle — as a black box, I would genuinely encourage you to crack it open. Upload your own documents. Make the AI explain the parts you have always nodded past. You do not have to become the expert. You just have to stop being a stranger to your own situation.
I would love to hear from anyone else doing this. Are you using AI to prep for your professional relationships? What worked, what surprised you, where did it fall short? Drop it in the comments — that is exactly the kind of conversation I learn the most from.
And if you want to talk through what this could look like for your own setup, reach out through the contact page. I am always up for it.
Comments
Leave a comment
AI SEO Is the New Visibility Game. Here's How I Picked a Tool For It.
People are asking AI instead of Googling. That quietly rewrites the rules of being found. Here's how I evaluated Profound versus Athena for work — and why we went with Athena.
How I Used Claude to Fight a $600 Insurance Denial — and Actually Filed a Regulator Complaint
A routine visit to a specialist turned into a billing mess and a denied claim. Most people give up at that point. AI is the reason I didn't — and why I filed a formal complaint with the state.
I Tried the Big AI Note-Takers. I Keep Coming Back to Granola.
I ran the major AI meeting-notes tools through real work — including a head-to-head with the Gemini note-taker built into Google Meet. One quietly won. Here is why Granola earned the spot.