It is being reported that on the U.S. West Coast, Big Tech is now going through large-scale workforce reductions, and systems integrator roles will be disrupted sooner or later by AI. However, when it comes to understanding client strategies, organizations, and operations, and to what the ideal system would entail, the most important systems development processes will still require the comprehensive thinking abilities of human beings, and while some tasks will be automated by AI, that does not mean that systems engineers will become obsolete. There are many operations in systems development that AI cannot handle, such as dealing with specification changes in the middle of development.
In the summer of 2025, I took part in a discussion with Stanford University’s Bio-Data Science Department on the use of AI in the healthcare field, and what impressed me was not only how AI can be used for drug development and treatment, but also how significant of a role AI is playing in making doctors’ clerical work more efficient as well. Capturing large numbers of images and scanning them for minute differences to detect incipient lesions is one of AI’s greatest strengths, but that does not mean that radiologists will find themselves irrelevant. This is because a radiologist’s job consists of some 30 different operations, and the diagnostic imaging handled by AI is merely one of that multitude (Professor Ajay Agrawal, University of Toronto).
iFLYTEK, a pioneering Chinese AI venture based in Hefei City, has launched an application that utilizes Spark LLM to create and score exams and to create lesson review materials, and the AI can even read student’s emotions from their facial expressions and use that data to provide individualized instruction. The aim here is to cut down on the instructors’ workloads and free up more time for interacting with students, and it seems that this technology has already been adopted by several thousand schools nationwide. And yet that does not mean the end of teaching as an occupation.

The arrival of AI agents has led intellectual work to be replaced at a rapid pace, and it is being reported that most head office operations will end up being replaced by AI, for instance, but the operations that human beings carry out are not purely standardized or routine. Through the exercising of our daily awareness, little improvements keep adding up, and workflows continue to evolve. Even if operations were replaced by AI, it would not be possible for the AI to autonomously identify problems, keep devising creative solutions, and make continual improvements. In other words, we should think of AI not as something that will replace humans, but rather as a tool for expanding human capabilities, and enhancing the productivity of human operations.
Having first taken off around 2017, DX has led to the creation of one new business model after another, for instance with telework and online meetings which changed the way people work, or with platform-type businesses like EC and Uber. Meanwhile, if AX only goes as far as replacing operations that people once did, and does not generate new business models, then the AI boom may eventually fade away, and it may take time before we see a return on the recent explosion in AI investments. That is to say, the critical point at issue will be whether AX is leading to business transformations.

There is a small bookseller in Sunagawa City in Hokkaido called Iwata Bookstore. The shop’s owner, Toru Iwata, began offering a service in 2007 called the “10,000 yen book selection” service. Customers fill out a book selection questionnaire with their personal histories and preferences, what they value in life, their worries, and so forth, and then send it to Iwata, who then selects his recommendations based on their answers and sends them the books along with a letter. This arrangement could even be called therapy through books.
Perhaps these kinds of new business models could be produced by AI. The more interactions AI has, the deeper its understanding of a client becomes and the more fully it can explore that client’s true needs. We call this “deep customization”, and if AI’s reasoning capabilities can be leveraged to ease its cognitive restraints when it comes to understanding clients, then perhaps it can generate new kinds of added value.

NRI first presented the findings of its research into “Digital Capitalism” at the NRI Dream Up the Future Forum in 2017, and published a book by the same name the following year. This publication focused on how DX will change people’s lives and our society. Now we see a shift from DX to AX, with AI continuing to make steady progress, and eventually these developments will create new services as well. At NRI, we will be carefully following these trends, and will continue to disseminate information about what the next social paradigm lead by AX will be.