On this site, I use the term "AIX" for rebuilding the role of the professional, and the relationship with the client, on the premise that AI exists. I call it "one step beyond DX," but I do not mean to claim technological superiority or the inevitability of progress. It is a word for thinking about who judges what, and how it is explained, once efficiency has advanced.
I do not believe a tax accountant's role is to hand over the answer AI produces as it is. It is to understand the circumstances and wishes of the person consulting me, to put into words the basis for a judgment and the parts that remain uncertain, and to support them until they can make their own choice with conviction. That, I believe, is the full extent of the role.
Below, I explain this thinking by drawing on answers from the 35-question Q&A.
Start from What the Person Wants
Section 01Tax and financial logic alone do not determine the answer for a given person. Faced with the same numbers, the right path differs from person to person. So before laying out options, I begin by understanding what that person wants.
The first thing I ask about is what the person actually wants. The tax figures and procedures can be worked out afterward in any number of ways. It is not unusual for people to be unclear about their own wishes, so we give them shape together as we talk. Once what they want comes into view, the right path usually narrows down on its own.
Sometimes I preface my advice with, "Speaking purely from the standpoint of money..." That is because what I can offer is only one view — the view from money. Everyone has things that matter more than money. So I hope people take my advice as one input to their decision and, in the end, measure it by their own yardstick. That is what I keep in mind when I speak.
I say clearly what the most reasonable approach is from a tax perspective. Whether to choose it, however, is the client's decision. The path with the lowest tax bill is not always the best path for that person. I lay out the advantages and disadvantages, and we look together for a form that fits what they want.
Delegate the Work, Keep the Judgment
Section 02Understanding people's wishes and explaining things so they can choose — this role cannot be measured by how much faster the work has become. On the contrary, it is precisely when AI changes how work is done that the questions arise: what should people check, and where should people decide?
I believe the tax accounting business, as it has traditionally worked, will no longer hold up. Being paid according to the volume of work cannot last once that work is automated. What will matter instead is what you checked, how you judged, and how you explained it. This is not limited to tax accountants; I think it is a change coming to all white-collar work.
Getting rid of laborious tasks is a good thing in itself. But I have doubts about whether that will translate directly into work that is easier.
Today's jobs involve a great deal of routine work, and there are actually not that many moments that require people to think for themselves. When the routine disappears and thinking becomes the center of the day, I suspect those who have had fewer chances to build that habit may find work harder, not easier. Whether the burden gets lighter depends not on the amount of work but on how its substance changes.
Repetitive work goes to AI. People check whether the results are correct, and decide what to hand over next. Broadly speaking, I believe these two will remain the human role. The more you delegate, the more the design of how you check and how you delegate becomes the substance of the work itself.
The groundwork for this way of thinking lies in my experience before becoming a tax accountant.
It gave me the habit of deciding on a goal and then thinking through the path to it rationally and thoroughly — questioning assumptions, checking with numbers, and rebuilding when something does not work. That way of working applies directly to tax work. I feel this way of thinking has been far more useful than any specific knowledge I picked up.
The other place that taught me the weight of judgment was the scene of disputes.
Being involved in tax and civil litigation made me realize how necessary a legal perspective is. Looking back once a dispute has arisen, it is clear how much the outcome depends on how records and documents were kept day to day. Since then I have been conscious of what to watch for and how to prepare in ordinary times. Preparation before anything happens is what works best.
No matter how much AI is used, the responsibility for professional judgment and advice cannot be transferred to AI. What information was assumed, what was verified, why a particular option is being presented — explaining what is uncertain as uncertain is also part of that responsibility.
Informed Choice and Professional Responsibility
Section 03Not only whether the work is finished, but how the person received the advice and whether they were able to choose for themselves — that is what I want to value. My own sense of having done good work also comes from there.
When clients are satisfied and also genuinely convinced. Satisfaction and conviction may look alike, but I think they are different things. Even with a good result, if people do not understand why it turned out that way, they cannot make the next decision on their own. The moment someone says "I see" after I have explained everything fully — that is when I feel it most.
This does not, however, mean guaranteeing the outcome someone hopes for. Nor does it mean taking on unlimited responsibility, or making business or life decisions on someone's behalf. Making professional judgments and taking on the responsibility of explaining them, while respecting the person's own choice — I want to hold on to both.