Do not mix up a crucial feature and a preference. ​

The word "интерес" in English is interest, which means "subject of commercial interest," "what interests us in the project." In system thinking, interest is simply historically referred to by a slightly different word: concern^[1], which is more accurately translated in Russian as "озабоченность": an important characteristic of the system, the subject of interest in the system. Subjects of interest are characteristics of the system that receive constant attention from various roles in the project; these are not role wishes, but rather where to move the subject of interest. The subject of interest for many roles and the interest of a particular role as a "role wish of where to move the subject of interest" — are different. The object of interest, a characteristic, is not "higher price" but "price," while interest/preference is exactly "higher price," where to move the subject of interest! For the price, there may be a different preference: a lower price, but the "object of interest"/"key characteristic" will remain the same—price.

Regarding the subjects of their interest, project roles, both external and internal team roles, constantly ask various questions. ISO 42010:2022 provides examples of such it should be organizational and departmental, not organizational and operations) ques- tions (remember that this is a standard on architectural documentation, but all other documentations will answer similarly formulated questions, only their subjects may be completely different):

• How is the system maintained?
• What system behavior is critical for security?
• Can the system-of-interest achieve compliance with mandatory national standards?
• What is the cost of operation?
• What risks, opportunities, satisfaction, survivability, harmonization, resource availability, complexity, and trust the given architecture offer?
• What is the transparency of distribution, as described in the standard model of open distributed processing^[2]?

The project team describes the system-of-interest and its environment, as well as the creators' systems in their complex graph so that both the project team roles and the external project roles have clear answers to these questions. Moreover, all involved roles take actions to realise their preferences on the interest topic (sometimes actors to achieve the preferences of some of their roles include the execution of work methods of other roles. Always remember that agents can show special creativity).

Roles are indeed concerned about a certain object; in their thinking, these labor/activity "concerns" prevail, and they act with the intention of realizing their preferences in the target characteristics. A seller will try to raise the price, a buyer— to lower it.

Each labor role, depending on its function/type of work, has one or more interests (subjects of interest with preferences in them)— while it is quite possible that different project roles/stakeholders have the same subjects of interest and the same or different preferences in them. This is very convenient: even if one role has two-three-five subjects of interest, the list of these important system characteristics will not be twice or five times longer than the list of project roles—the list will be much shorter. If you need toprepare materials/descriptions/models for fair trade-offs in interests, their numberwill be proportional to the number of important characteristics, not the number ofpreferences, not the number of project roles. Further, the preferences of a largenumber of roles will be negotiated through a few descriptions reflecting a smallnumber of subjects of interest.

The functional scheme of an electric scooter as a description reflecting some of its important characteristics will be used by many roles – both those advocating for higher power/performance: characteristic and those advocating for limited energy consumption: characteristic and those advocating for maximum speed: characteristic, and those concerned about its charging speed. This description/model allows for negotiating multiple preferences for the various characteristics of the electric scooter as subjects of interest of external and team roles in the electric scooter development project.

Even with the same subjects of interest/important characteristics, preferencesdifferent internal/team roles and external roles may vary significantly in the meaning ofthese characteristics. If roles such as "safety officer" and "salesperson" meet, then the subject of interest for the electric scooter will definitely be "maximum speed," but their preferences/evaluations of interest will differ: for one, to achieve maximum safety, the maximum speed should be lowered as much as possible, and for the other, it should be increased to persuade purchases and have a speed advantage over competitor scooters. Of course, the appearance of a pair of roles with such different preferences for one interest is not necessary, but such pairs are quite common, not only in terms of "cost" or "maximum speed" interests.

Negotiations on the values of important characteristics (sometimes called productive conflict) go with any subjects of interest. For example, the characteristic "weight of the electric scooter" has a preference for "less" from the consumer, but for the designer, it may have an interest in "more weight" to accommodate a larger battery or use a more durable and therefore heavier design. The designer may acknowledge that "more weight" is bad, but it is a "necessary evil" for him, it is good! If the decision is left to the designer, the scooter may end up being excessively heavy for the consumer, even though it looks excellent based on many other characteristics. It is better to defend the consumer's interest in weight: characteristic by the marketer/salesperson as a team member than to accept the "balanced decision" made by the designer.

Remember that if both project roles with conflicting preferences on one characteristic are played by one actor/role player, and the productive conflict occurs in one head, then it is called a conflict of interests^[3]. It is useful to separate conflicts of interests among different people/actors: when there is a productive conflict, roles monitor each other's mistakes, but when this error control occurs within one person's head, it is difficult: a performer might favor the decision of one of the roles, even if that decision is erroneous. By separating the conflict of interests into roles for different actors, the discussion on preferences in one subject of interest is public; hence, errors in arguments can be pointed out by some third parties who can also provide solution suggestions for the productive conflict resolution; decision-making can be documented and verified by independent experts. Thus, in the absence of a conflict of interests (a productive conflict within one actor's head) and by converting the conflict of interests into a productive conflict, more quality reconciliations of preferences within the productive conflict in different heads of different actors can be carried out, leading to fewer errors, and the different role preferences will be accounted for more qualitatively, making the system more successful.

However, if you try to separate these different conflicts of interests between different heads (if they are people, or even if they are departments) literally with all the identified conflicts, nothing will work: there are a huge number of conflicts of interests, and even considering that one head can participate in multiple conflicts of interests, you still need to make sure that roles engaging in intensive negotiations are organizationally closer to each other— negotiations should not be among all with all but within some organizational units (this will be discussed later: good modularity is important both in target systems-products and in the organizations of creators), so, in the end—inside one agent’s head.