紳士淑女録 #003｜水の鬼平、原理原則を守る——山田正幸とオルガノの知の継承
I affectionately call Masayuki Yamada “Mizu no Onihei” and “Mizu no Ojiki.” These are personal nicknames expressing respect for a leader who wants to preserve engine
I affectionately call Masayuki Yamada “Mizu no Onihei” and “Mizu no Ojiki.” These are personal nicknames expressing respect for a leader who wants to preserve engineers’ knowledge and technical discipline while changing how work gets done. They are not his official titles. This portrait comes from reading, not a personal meeting.
What stayed with me after the Forbes JAPAN interview was a demanding idea: AI may help somebody perform a task, but understanding the principles behind that task still matters. The organizational challenge is to make expertise reusable without making the next generation dependent on answers it cannot explain.
A career worth looking at
Yamada completed graduate studies in botany at Kyoto University in 1985, according to the Forbes profile, and joined the company now known as Tosoh. Organo’s official biography records leadership roles in Tosoh’s bioscience businesses, followed by responsibility across functional products, research planning and engineering. He became an Organo director in 2019, senior managing executive officer in 2021, and president in June 2022. [1,2]
I would like to ask him how that experience influenced his approach to cultivating specialists. The connection is a question, rather than an explanation I can establish from a résumé.
Water as an operating capability
Organo’s technical materials describe combinations of separation and purification methods, including membranes, ion exchange and ultraviolet treatment. The practical question is how a system meets a customer’s contamination requirements under its particular conditions. A purity headline alone cannot capture that work. [3]
My analytical hypothesis is that switching friction comes from several capabilities working together: design, analytical verification, commissioning, troubleshooting and accumulated knowledge of a particular factory. Installed equipment is only part of that relationship. This needs to be tested through customer retention, service performance and project economics, rather than assumed from the industry’s importance.
Two different AI projects
The Forbes account describes both knowledge-transfer work with LIGHTz and an internal generative-AI platform. LIGHTz’s own announcement explains its role in structuring experienced specialists’ tacit knowledge. These initiatives address related but distinct problems: retaining technical reasoning and improving everyday work. [1,4]
The interview’s figures need careful labels. Reducing some tasks to less than one-tenth of their former duration is a prospective possibility, not a companywide result. The reported seven-month investment payback is a theoretical calculation using labor-cost equivalents of time saved. It does not establish an equivalent reduction in cash expenses. An internal AI platform also does not mean Organo developed its own foundation model.
My questions are about deployment quality: are answers traceable to approved sources, are exceptions preserved, and can junior engineers explain why a recommendation is appropriate? These are research questions, not claims about controls already implemented.
Revenue is not the same thing as cash generation
Organo’s FY2026 materials distinguish plant, solution and functional-product activities. Solution revenue was approximately half of water-engineering revenue, but roughly 43% of consolidated revenue; the denominator matters. Its solution business includes maintenance and improvement services, while equipment-owned services introduce financing and recovery questions. [5,7]
A particularly useful disclosure appears on page 12 of the May 2026 results presentation. Investment in equipment-owned service projects is included in operating cash flow in the cash-flow statement; the presentation reclassifies it into investing cash flow for its analytical chart. Comparing the chart with reported operating cash flow without reconciling that treatment can mislead. [5]
My interpretation is that durable customer relationships can coexist with substantial capital commitments. I want to see contract duration, cash recovery, financing costs and asset risk alongside recurring revenue.
The wider industry is a test of the thesis
Kurita’s own recruitment materials describe ultrapure-water supply, operating optimization, and the transfer of veteran knowledge to younger staff. Those challenges are therefore not unique to Organo. Nomura Micro Science’s FY2026 release offers another useful perspective: total sales fell while maintenance and consumables grew, with large-project timing affecting the result. Industry importance does not guarantee uniform company performance. [8,9]
Customers are also working to use water more efficiently. TSMC’s September 2026 account describes nanobubble cleaning and monitored process control, with further water savings projected from wider deployment. That example is not evidence of an Organo contract. It shows why I must study water intensity and customer economics alongside fab expansion. [10]
What I would ask Yamada
How does Organo distinguish knowing an answer from understanding a process? How are unusual cases and design exceptions preserved in its knowledge system? What evidence will show that time saved becomes additional engineering capacity, better quality or stronger returns?
My admiration is for the responsibility implied by those questions. “Mizu no Ojiki” means, for me, somebody who takes the work and the people behind it seriously. I want to keep reading the disclosures with that same seriousness.
Sources [1–10] and related research are listed below. Japanese sections contain the detailed financial tables and research questions.
