How to Behave Right in the Wrong System

Adorno’s phrase that there is “no right life in the wrong one” captures a contradiction that reaches far beyond AI. We are asked to make the right decisions within conditions we did not create and often cannot leave behind.

This is especially visible in AI. Companies decide which systems they build, which data they use, under what labor conditions they are produced, and which social and ecological costs they are willing to accept. Only afterward does the debate begin about how we should use these systems responsibly. Check the output. Do not enter sensitive data. Watch for bias. Think about resource use. Consider carefully when the use is still defensible.

Responsibility moves downward while power over the conditions remains largely at the top. That does not make individual decisions meaningless. People can set boundaries, refuse, boycott, and avoid harm. The problem begins when these decisions are treated as if social problems could be solved by enough individual caution. Then politics becomes a question of personal morality and structural change becomes a search for the most correct way to behave.

Luise Freese speaks about this trap, about guilt and complicity, and about the understandable desire to act at least correctly within contradictory conditions. She also asks what is lost when we spend so much energy optimizing our own participation instead of changing the conditions under which we make decisions at all.

A(I)mbivalence therefore offers no guide to living morally impeccably in the wrong system. It is interested in the point at which personal responsibility is no longer enough and the political question begins.