guest ::
| Home > Publications database > Optimal Decisions and Why Not to Ask AI for Them |
| Home > Publications database > Optimal Decisions and Why Not to Ask AI for Them |
| Contribution to a book | FZJ-2026-00911 |
2025
Springer Nature
Switzerland
Abstract: Although we make decisions all the time in our own personal lives and as scientists, there is still a surprising amount of ambiguity about what constitutes an optimal decision. We recently started to improve or fully outsource some of our decision-making to AI technologies that are built to provide recommendations or decision-making support. In a society that relies more and more on such support systems, it is also remarkably unclear how optimal decision-making could be further boosted. Although AI outputs often successfully assist human decision-making processes, what has yet to find sufficient attention in the debate is an encompassing understanding of the conditions for optimal decision-making with and without AI. This chapter (i) critically investigates the understanding of and overall pursuit to make optimal decisions, and (ii) describes the discrepancy between human practice of decision-making and the so-called AI decision-making processes. The concept of decision-making is connected to concepts of agency and thus part of a normative framework. One result of this investigation is that while AI can provide some calculative assistance to find a defined optimum, it cannot make optimal decisions.T
|
The record appears in these collections: |