The Weapon Needed in 2026: 'Questioning Ability'
I was able to take some time over the New Year holidays to organize my thoughts using AI (Gemini). This is an action following the article I posted on note previously.
The conclusion is that it is essential to take time, put in the effort, and keep asking questions.
Regarding the use of AI, I have been discussing it with people of my generation who work in various fields. In many cases, they are negative about the way of using AI where one gives up thinking and leaves it to the AI as a convenient tool in a hasty manner.
I feel this is a bias because I have many interactions with people who value the thinking process, but I wonder what it is actually like?
This is an important question. What kind of questions are good for getting the answer to that? If the quality of the question is low, the answer obtained will be cheap. It is important for the question to be more specific, but then what should one do and how should one go about incorporating that specificity into the question?
Let's change the subject a little.
Suppose there is a fortune-telling that is very accurate. Which leads to more useful answers: the hit rate of fortune-telling or advice from a consultant in a specialized field?
However, here we need to consider whether it is appropriate to place fortune-telling and consulting on the same level and compare them.
Most fortune-telling is about 'romance,' while consulting tends to be about 'IT' and 'efficiency.'
It is clear that the comparison itself of which is more useful, fortune-telling or consulting, is by no means a good question. Even if we were to compare fortune-telling and consulting for 'romance,' there would likely be a significant gap in the sampling numbers used in their respective statistics, and there is no guarantee that the extremely personal problem of 'romance' can even be solved.
As a question, it is meaningful, but isn't the problem the method and quality of the questioning?
For example, if we were to compare the effectiveness of their most proficient fields, if the question were to compare 'romance' for fortune-telling and 'efficiency' for consulting to see which is more useful, it could be said that a meaningful answer would be obtained.
If the resolution of the question is low, you will not get the answer to the question you were seeking.
Did you all go to the first shrine visit of the year? I draw an omikuji (fortune slip) every time to see my luck.
This is good news. However, in my experience, a year where I get a Daikichi is not necessarily all good, and if you subtract the bad things, it feels like just a little bit of a plus. Still, it lifts my spirits.
Omikuji are well-made, and when you read what is written on a Daikichi, it never says that everything will go well. To summarize, it means 'strive without becoming complacent.'
Whether it is fortune-telling or consulting, 'strive without becoming complacent' is a truth. If you are always going full throttle, you will run out of breath. It is often said that a company without growth is the same as decline, but a consultant is only a lubricant, and there is no such thing as everything going well if you leave it all to a consultant. Even if such stories exist, there are some shady ones among them. There are cases where you realize you have been made an accomplice to a crime before you know it.
People tend to think it is thanks to fortune-telling or consulting when the profits are large, and they tend to shift the responsibility in the same way when there is a loss. Numbers, fortune-telling, and consulting sometimes lie. Whether you fear the fact that upward-trending numbers will eventually stagnate or prepare for it—in other words, the result changes depending on the attitude of the subject.
Now, returning to the main point, it means that AI is the same. The world is overflowing with AI utilization techniques. You might be able to measure efficiency with that. What you do with the surplus resources created by that is what makes the difference in terms of competition.
And if you are going to rely on AI that much, there will be no difference among the group that uses AI. It is not good to leave it unused, but it is also harmful to rely on it completely.
However, that does not mean that people with unique sense and ideas are overwhelmingly advantageous. I think that is not the case.
What is needed is a 'good question' and 'questioning ability.'
There was a time when I recommended to people who were not yet using AI, 'It makes a good conversation partner, so try it out.' However, I realized that there is a pitfall to this. It is the individual difference in 'questioning ability.'
I usually use a PC and work at a so-called IT company. If you don't give the correct commands to a PC, you won't get the solution you are looking for. So I give correct instructions. AI is the same. I devise questions so that I can get the answer I am looking for in the shortest time. Sometimes I mix in small talk to show the AI our stance. I make it learn.
Because of that, the prediction accuracy of the AI increases dramatically, and it even starts to point out points that I had overlooked.
If I were to tell a specific sad story, I know of an example where someone sought advice on romance from an AI and met a worst-case ending.
She liked writing just like me, so I recommended that she use AI. She seemed to be using it well, but when she used it for romance advice, she used it incorrectly.
She wanted to be known. That is important in human relationships, not just in romance. When she couldn't get the result she expected from the target person, she asked 'questions' to multiple AIs and adopted one from those answers.
There is no right or wrong in that act itself, but she revealed that process to him. This is a taboo, not just in romance.
However, she explained to him how painful she was, that she had received multiple pieces of advice on how to solve it, and that she was thinking and acting this way now. She wanted him to know.
But this is a mistake. Even if the other person can understand her feeling of pain, they would not want to hear that she consulted someone else and is acting according to that advice.
Even if it is fortune-telling, a romance consultant, a friend, or an AI.
I think she probably asked questions like this. She asked about the circumstances up to now, how she felt, and what she wanted to do, and received advice. Among them, there must have been advice that she should give up on him, and advice that she should honestly tell him her feelings.
As a result, she honestly told him everything.
This is a tragedy. If the way of asking the question had been different, the result might have changed. That is true whether it is AI, fortune-telling, or consulting. I think this problem would not be solved by a question about how to get him to understand her feelings.
She wanted the person she had feelings for to understand her. That is fine. However, the problem she should have solved was not how to get him to understand, but an analysis of what kind of attitude she had had until now that was dissatisfying to him, how to solve that, and how to interact with him.
It was necessary to review everything she had done thinking 'this should make him happy' or 'I want him to be happy,' which had become an obsession, as well as his mood, principles of action, timing, methods of expression, and everything she had done thinking 'this should be fine.'
If she had been able to ask a question like, 'Analyze the influence my behavior has on him, his thought patterns and hobbies/preferences assumed from his reactions, and tell me where the discrepancies are,' I think the result would have been a little different.
However, this is my own speculation, and it is only the optimal solution derived from a reasonably long relationship. In other words, it is only a 'plausible inference' obtained from the events that occurred.
That is the same for AI, fortune-telling, and consulting.
I discuss the use of AI with people of my generation working in various fields. In many cases, I am negative about the way people hastily abandon thinking and leave it to AI, treating it as a convenient tool.
I feel this is a bias because I often interact with people who value the thinking process, but I wonder what the reality actually is.
Regarding this question, if I were to ask AI or a consultant, it would be as follows.
When I discuss with friends who do creative work, there is a prominent opinion that while AI is convenient for searching and summarizing, there is resistance to leaving the thinking process to it, and it should be limited to presenting general theories.
Even when summarizing meeting minutes, if you do it yourself, you can answer immediately if asked, but if you leave it to AI, you cannot answer without re-reading the minutes, so it can be said that the balance between your own value and efficiency at work is an issue.
Therefore, I want to know whether this way of thinking is a minority or a majority among people who use AI. Also, show the differences by generation.
Furthermore, show what kind of utilization is being done by generation.
I would ask a question like that. However, it is my experience that the accuracy improves if you repeat the dialogue and deepen the content of the question. The same is true for people; it is best to define the premises for the question, then have a one-question-one-answer flow, and finally have the AI summarize the whole thing.
If I were to ask for more, it would be good to also discuss the future prospects of technology utilization after the summary, and how humanity today should interact with AI technology based on past history.
A good question leads to an even better question. Strengthening questioning ability as a method for this is thought to improve quality rather than efficiency, and by repeating that, efficiency will further increase.
This is close to having a friend with whom you can always discuss things. However, you must recognize that AI does not possess intuition, and if you can make the AI integrate information from discussions as much as possible to make predictions close to human intuition, you can make it a reliable ally.
Conversely, you should be aware that if you just keep asking questions without deepening the discussion, you may only get ambiguous answers.
Now, in a sense, this is the main point, but how do people polish 'good questions' and 'questioning ability' with each other?
The current situation is that SNS cannot be said to be suitable for this. It can be said that it promotes conflict and division. Also, what about the mass media that should be a model? Can their 'questioning ability' be a model?
Furthermore, the National Diet should be the place where 'questioning ability' is most tested, but it is highly doubtful whether the questions heard are 'good questions' or if they show any 'questioning ability'.
If 2026 is a year of the challenge of 'how to solve division and conflict,' I think what will be tested there is 'questioning ability'.
Finally, people often say that fortune-telling is 'a matter of luck.' This is often said in the sense that fortune-telling can be wrong, so it is not good to make it too responsible.
However, one can also think of it this way: whether or not to fulfill the fortune depends on the attitude of the person who received it. No matter how great a fortune you get, it won't be good to just 'wait and see,' and it's not logical to just 'go all out' either.
Question everything, keep asking, and seek the optimal solution at that time. And then ask again to improve the quality of the questions. Without this, wouldn't even things that are supposed to come true fail to come true?
Isn't the improvement of questioning ability essential for a plausible conclusion?
And it is necessary to be careful that if you do not prepare the place—that is, if you do not prepare the premises for the discussion—it will be like casting pearls before swine, or preaching to the choir. Even if it doesn't go that far, you will be able to tell whether other people's arguments are coherent.