Escrever uma avaliação PREreview

This article examines a central conceptual problem in debates on radical human enhancement: when does a technologically produced increase in a human capacity count as a genuine enhancement rather than merely an augmentation? It argues that radical transhumanism faces an underexamined inferential difficulty. The successful optimization of a measurable or technologically tractable capacity—such as cognition, memory, longevity, endurance, or affective control—does not by itself establish that the resulting state is better for the person in an all-things-considered sense.

Drawing on Goodhart-type failures of proxy optimization, the article develops the concept of Anthropological Goodhart Vulnerability to explain how a locally valuable capacity can become an unreliable guide to broader human improvement when subjected to radical optimization. It identifies three mechanisms through which this can occur: extremal extrapolation beyond familiar human conditions, target substitution, and the suppression of trade-offs with other dimensions of value. A minimal formal framework is used to distinguish operational success from normative enhancement and to show why an additional normative bridge is required before increases in a targeted capacity can justify claims of overall improvement.

The analysis is further extended through the concept of Reflexive Anthropological Goodharting. In radical forms of self-transformation, an intervention may alter not only the capacity being enhanced but also the preferences, evaluative standards, and conception of flourishing through which the transformed state is subsequently assessed. This creates a risk of endogenous self-validation, particularly when post-intervention endorsement is treated as sufficient evidence that the transformation was beneficial. The article also develops a Local-to-Global Enhancement Problem, showing why a sequence of individually accepted and apparently beneficial transformations does not necessarily establish that the overall trajectory constitutes continuous improvement.

To address these problems without adopting a bioconservative position, the article proposes a Non-Self-Certification Principle and a set of Provisional Enhancement Conditions. These conditions emphasize independent normative support, sensitivity to multiple dimensions of value and possible trade-offs, cross-perspectival justification, revisability, and a justificatory burden proportionate to the radicality and irreversibility of the intervention. The overall argument does not reject technological transformation or posthuman possibilities. Instead, it establishes an epistemic and normative constraint on claims of radical enhancement: technical success alone cannot certify that a technologically produced transformation constitutes genuine human improvement.