A Decision Theoretic Perspective on Artificial Superintelligence: Coping with Missing Data Problems in Prediction and Treatment Choice

2025-09-15

AI Summary

This paper addresses a fundamental bottleneck in artificial superintelligence (ASI) development—the identification problem arising from missing data: a structural uncertainty under which causal inference remains unreliable regardless of sample size. This limitation explains why current machine learning–based AI systems fail to consistently outperform human judgment in prediction and treatment selection. Method: The paper introduces a decision-theoretic framework, formalizing identification feasibility as a novel benchmark for intelligence. Integrating econometric identification analysis and causal inference techniques, it models intervention effect estimation under missing-data regimes. Contribution: It demonstrates that mainstream ML architectures intrinsically lack mechanisms to represent or reason about identification uncertainty. The paper proposes a paradigm shift beyond statistical fitting—advocating ASI designs that explicitly incorporate counterfactual reasoning and decision-theoretic robustness to identification ambiguity. This reframes ASI advancement as fundamentally requiring causal epistemic competence, not merely predictive accuracy.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Enormous attention and resources are being devoted to the quest for artificial general intelligence and, even more ambitiously, artificial superintelligence. We wonder about the implications for our methodological research, which aims to help decision makers cope with what econometricians call identification problems, inferential problems in empirical research that do not diminish as sample size grows. Of particular concern are missing data problems in prediction and treatment choice. Essentially all data collection intended to inform decision making is subject to missing data, which gives rise to identification problems. Thus far, we see no indication that the current dominant architecture of machine learning (ML)-based artificial intelligence (AI) systems will outperform humans in this context. In this paper, we explain why we have reached this conclusion and why we see the missing data problem as a cautionary case study in the quest for superintelligence more generally. We first discuss the concept of intelligence, before presenting a decision-theoretic perspective that formalizes the connection between intelligence and identification problems. We next apply this perspective to two leading cases of missing data problems. Then we explain why we are skeptical that AI research is currently on a path toward machines doing better than humans at solving these identification problems.

Problem

• Addressing missing data problems in prediction and treatment choice
• Examining AI's limitations in solving identification problems with missing data
• Assessing machine learning capabilities versus human decision-making under uncertainty

Innovation

• Decision-theoretic perspective on identification problems
• Addressing missing data in prediction and treatment
• Skepticism on current AI outperforming humans