Who Decided That Was the Problem?

Part 25 of 30 · Series date: July 5, 2026

A person is lonely. One proposed solution is an artificial companion. Another is transport the person can actually use. A third is a community space where they can meet people. Each proposal may help, but each describes the underlying problem differently.

Before asking whether a technology solves a problem, we should look at how the problem was defined. That definition often decides which facts count, which choices we see, and whose duties remain in view.

Solutions contain assumptions

If loneliness is defined as the absence of comforting conversation, a conversational tool may look sufficient. If loneliness means being shut out of two-way relationships, the tool may meet only part of the need. Neither description should be imposed without listening to the person concerned.

This does not mean there is one perfect explanation for every difficulty. Problems can have several causes and require several responses. A useful tool need not solve everything to be worth using. The danger comes when a partial answer is sold as complete because it fits the product.

The strongest defense of practical technology is that people need help now. Waiting for ideal social reform can leave them unsupported. The strongest reply is that immediate help should not erase the larger duty. We can relieve a burden while still changing the conditions that produce it.

What the measurement misses

Suppose a hypothetical ministry evaluates success only by the number of meals served. The count matters. But it may miss respect, access, trust, or whether people can return for help. A system aimed at that count might raise the number while weakening the personal encounter.

The solution is not to abandon measurement. Without facts, a group can hide failure behind its good aims. The solution is to remember what the measure represents and what it leaves out. A number is evidence about an aspect of the work, not a complete account of its worth.

Similarly, a school might distinguish performance on a task from learning that transfers to a new situation. An employer might distinguish output from lasting work. A healthcare organization might distinguish throughput from the quality of care. The relevant measures should follow the purpose rather than quietly redefine it.

Who has a voice?

Problem definition is also a question of power. The person purchasing a system may experience a different problem from the person subjected to it. An administrator may want fewer appeals; an applicant may need a fairer decision. Optimizing one goal could worsen the other.

The people affected need a say early enough to shape the goal. They should not just react to a finished product. Their experience is essential, though it will not always produce agreement. Different people may have good reasons to want different results.

For Christian readers, love of neighbor can discipline this process. It asks us to encounter a person rather than reduce them to an inconvenience to be managed. For readers with other foundations, respect and justice can support a similar insistence on involvement.

A solution can answer the purchaser's problem

Imagine a service getting many calls because people do not understand its decisions. Management defines the problem as call volume. Applicants define it as confusion and lack of a way to appeal. Making contact harder might reduce calls. It would meet the manager's target while making the applicant's problem worse. Nothing in the efficiency metric itself forces the organization to notice the difference.

This is why problem definition is already an exercise of power. Whoever holds the budget may decide what counts as the problem. Other people's suffering may be pushed into the background. Affected people should take part before the chosen measure has hardened into the objective. Consultation after the system is built may leave them discussing details of a goal they never accepted.

Their involvement will not always produce agreement. Some people may prefer speed, others explanation, and others personal contact. Good design must handle these tensions rather than claim that listening produces one obvious answer. Make the tradeoffs clear and answer for them. Do not pretend that every group has just one view.

Measures are useful because they leave things out

A measure simplifies. That is its strength: it allows a pattern to be seen. It is also its limit. Counting meals can reveal whether a kitchen is meeting demand. It cannot tell us whether guests are treated with respect or get the food they need. It cannot tell us who is still hungry outside because they cannot enter.

The answer is not to abandon numbers in favor of reassuring stories. Stories can also be selective. The stronger practice uses several kinds of evidence and asks what each can miss. A count, a direct observation, an appeal, and a talk with a person can each reveal different things. None should be allowed to become a complete replacement for the purpose.

This matters when a system is built to pursue a chosen target. It may pursue the target very steadily. If the measure is incomplete, consistent pursuit may magnify the omission. The group must notice when the number stops tracking the good it was chosen to measure.

Ask what would happen without the product

A fair comparison includes a plausible alternative. Suppose a community plans to provide artificial companions. What else could it do with the same money? It might improve transport, open accessible meeting places, or support caregivers. The answer may still favor the tool, or a combination. But without the comparison, the product's existence can make its own approach seem certain.

The alternative should not be an imaginary perfect society. People need help under actual limits. A technology can provide a valuable partial response while larger reforms remain unfinished. The wrong comes when partial help is sold as complete. It is also wrong to use it to end duties it does not meet.

Before adopting a system, write a sentence describing the desired human change without naming the product. Aim for “people understand and can question their help,” rather than just “fewer calls.” For students, ask whether they can explain a new problem. For loneliness, ask about dependable connection, not just longer chats. These words do not solve the problem. They help reveal whether the service helps the person or just makes the person easier to manage.

A better first question

Before adopting a tool, state the problem in several ways. Ask what would count as success from the view of each affected group. Consider at least one way to change the setting rather than the person. Then look at whether the proposed technology adds value within that broader understanding.

This approach may still lead to enthusiastic adoption. It may also reveal that a simpler intervention is better. Either result is better than letting a product define what a person needs. The most serious act of judgment may occur before anyone presses the button.