The Wrong Axis and the Tower Will Fall A Blueprint Starting with the Content of Ethics December 2025

The Wrong Axis and the Tower Will Fall A Blueprint Starting with the Content of Ethics December 2025
No matter how elaborate a system or framework is prepared, if the content of the ethics that serves as its foundation is distorted, what is created will be nothing more than a refined hazard. To decide on the structure before the content is akin to creating a device that neatly amplifies wrong values. Therefore, there is a position that what should be asked first is the content of ethics: what kind of world is good, what should be protected and what should never be allowed, and only after that should we finally consider institutional design and technical frameworks.
Looking back at history, there are many examples of organizations and laws that have only been formalized, but have justified discrimination, surveillance, and violence. Proceeding with organizational structure and algorithm design without adequately examining the content of ethics is a dangerous shortcut that mechanizes value judgments and spreads false assumptions to every corner of society. Especially with a technology such as AI, which, once introduced, will have far-reaching effects, the principles and values on which the structure is premised must first be examined to ensure that they are consistent with respect for human rights and fairness.
International AI principles also reflect this content-first mindset: the OECD AI Principles take values such as human rights, democratic values, equity, and transparency as a starting point, followed by a policy and risk management framework; the EU AI Act also sets out the objectives of protecting health, safety, and fundamental human rights first, and The EU's AI Act also sets the objectives of protecting health, safety, and fundamental human rights first, and then designs a structure of risk-based regulation to realize the objectives. In Japan, too, human-centered AI social principles are established first, and governance guidelines are developed later as implementation guidelines.
If sophisticated algorithms and elaborate governance structures are built on top of the distorted content of maximizing organizational profits and maintaining control of a particular group, it will become a dangerous device with only the outward appearance of being in place. This is why the order in which questions are asked is so important. What should be decided first is what ethical content is shared and whether the institutions and technology are designed in a way that does not betray that content. Starting with the soundness of the content, not the cleverness of the structure, will be the condition for the coming AI era.