AI is advancing so rapidly that the idea of machines taking human jobs no longer feels like a distant prediction. What surprises me most is that many of the jobs now under pressure are not manual ones, but highly skilled intellectual professions—doctors, programmers, scientists, and perhaps teachers. As a pathologist who uses AI almost every day, I increasingly wonder not whether AI will change my profession, but how much of it will eventually remain in human hands.
Recently, I have not been hearing very much about political developments.
The news seems to be filled day after day with floods and typhoons.
Of course, that is not really the case. There are also reports here and there about the consumption tax, U.S.-China relations, Ukraine, Iran, and other issues.
Japan’s extraordinary session of the Diet is about to convene, so I expect various political debates to begin. Yet along with an increasingly complicated international situation, rapid advances in AI are bringing fundamental changes to the structure of society.
The future looks increasingly chaotic.
U.S. President Donald Trump apparently wants AI to be called SI, or “super intelligence,” from now on, and promoted the term during his recent meeting with Chinese President Xi Jinping.
Perhaps he hopes to go down in history as the man who named it.
But whether we call it AI or SI does not fundamentally change what we use it for, so changing the name hardly seems to be the essential issue.
If OpenAI changed its name to OpenSI, though, things might begin to feel a little different.
Recently, I have begun to understand much more clearly what it really means when people say that AI will take human jobs.
I had not expected it to take over so much of the work done by intellectual professionals.
Even looking briefly around the medical profession, there are already situations in which AI can outperform physicians in diagnosis.
For an internist, if a patient’s condition, chief complaints, and clinical data are entered into an AI system, it can already produce highly accurate differential diagnoses for many diseases and even suggest treatment strategies.
Surgeons, too, should be able to reduce their workload considerably by using AI to analyze radiological images and assist them in deciding how to proceed.
As for pathologists, if we prepare the specimens correctly and accurately identify the histological findings, we can then discuss those findings with AI and work our way toward a diagnosis.
But even pathologists have no idea when “AI unemployment” might come knocking at our door.
In all these cases, however, human involvement is still assumed.
So the future may depend on how much meaningful human involvement remains possible.
And I have another concern.
Will the coming generation of doctors, who grow up practicing medicine with AI from the beginning, still have enough knowledge and experience to check whether AI is right?
Programming once seemed to be one of the professions of the future, and many people became programmers during its golden age.
Now, with the extraordinarily rapid development of AI, programmers themselves seem to be facing difficult times.
I recently read in a newspaper about AI solving a difficult mathematical problem.
Scientists, too, may eventually find themselves under considerable pressure from AI.
Perhaps the idea that truth itself will be revealed by AI will eventually apply to every field.
Education may also increasingly be handled by AI.
Teachers might someday find that their main remaining role is to interact with students as older and more experienced human beings.
Yet when I hear recent reports of disgraceful misconduct by some teachers, I can also understand why parents might sometimes feel safer entrusting certain aspects of education to AI.
Perhaps children would need to interact closely only with carefully chosen adults.
And for children who find group activities particularly difficult, AI might sometimes be a better partner for learning.
Some people will probably resist and refuse to use AI.
But they will find it difficult to compete with people who do use it.
Even if individuals resist, they are unlikely to win against groups in which the use of AI is simply taken for granted.
If so, the number of occupations that disappear or shrink because of AI may be far greater than we currently imagine.
For many people, it may soon be a case of, “I could be next.”
And remarkably, many of those people will be intellectual professionals.
I still do not know whether all of this will ultimately be good or bad for humanity.
I do not even know whether the answer will become clear while I am still alive.
But the world as it exists when I die will be the final point of the future that I personally get to see.
So I want to keep thinking carefully about what that world will look like.
The future may belong not simply to those who use AI, but to those who can still judge whether AI is right.
・・・AI-driven job loss — AIによる雇用喪失
AI-driven は「AIによって引き起こされる」という意味。
intellectual professional — 知的専門職
医師、研究者、法律家など、専門知識を中心に仕事をする人。
differential diagnosis — 鑑別診断
可能性のある疾患を複数挙げ、絞り込んでいく医学用語。
human involvement — 人間の関与
AIに完全に任せず、人間が判断や確認に参加すること。
come knocking at one’s door — 身近な問題として迫ってくる
直訳は「戸口をノックしに来る」。ここではAI失業が自分にも迫るという比喩。
take something for granted — 当然のことと考える
本文では、AIを使うことが最初から当然になっている状態。
be under pressure — 苦境に立つ、圧力を受ける
職業や業界が変化を迫られる場合にもよく使う。
I could be next. — 次は自分かもしれない
日本語の「明日は我が身」に近い自然な表現。
#ArtificialIntelligence #AIJobs #FutureOfWork #MedicalAI #Pathology #Kamakura #JapanesePathologist
















