Position: Machine Learning Research Should Be Guided by Explicit, Pluralistic Models of Human Purpose

Utsav Gupta

Abstract: Machine learning systems increasingly shape attention, work, education, and social life, yet ML research often treats the question "what is this for?" as external, relying on proxies such as accuracy, engagement, or preference satisfaction. This position paper argues that ML research should be guided by explicit, pluralistic models of human purpose, understood as supporting people's capacity to pursue meaningful, self-chosen life projects with agency. The paper proposes three community practices: (i) purpose articulation, a structured "Purpose Statement" that specifies intended beneficiaries, mechanisms, and falsifiable failure modes; (ii) purpose evaluation, which measures impacts on agency and meaning alongside task performance and harm; and (iii) purpose governance, which updates purpose frameworks through transparent, participatory processes to reduce unaccountable value-setting. This framing enables concrete technical research directions, including objective design beyond preference satisfaction, benchmarks for agency and meaning, pluralistic system behavior, and institution-aware alignment. The paper provides stakeholder-differentiated recommendations for researchers, benchmark creators, conference organizers, and funders, and addresses credible objections including value neutrality, feasibility and measurement validity, the claim that harm prevention is sufficient, and risks of ideological capture or paternalism.

Lay Summary:

Machine learning shapes what people read, decide, and do, yet ML research rarely asks "what is this system for?", instead optimizing proxies like accuracy, engagement, or user preferences. The catch is that those proxies reshape people: a tutoring system that maximizes session length can leave students less independent, and an AI writing tool that boosts productivity can shrink the writer's voice. What gets displaced — agency, skill, and meaning — is the harm.

This paper argues ML researchers should make human purpose a central research concern, and proposes three practices. First, a short "Purpose Statement" for papers claiming real-world relevance, stating who the system should help, what could go wrong (dependency, skill loss, manipulation), and what would count against it. Second, evaluations that measure whether systems support people's agency and capacity for meaningful action, not only task performance. Third, governance through proposed public registries and external review, so a system's claimed purpose can be challenged rather than buried in private choices.

None of this requires agreement on a single theory of the good life. It asks researchers to state what they're building for, measure whether they're achieving it, and let others contest the answer.

Primary Area: Research Priorities, Methodology, and Evaluation

Keywords: human purpose, AI alignment, pluralistic alignment, human flourishing, agency, evaluation methodology, research norms

Submission Number: 978