After Orthogonality: Virtue-Ethical Agency and AI Alignment
Philosophical reframing of AI alignment has long-run implications for how safety standards and governance frameworks conceptualise AI intent and values.
Key points
- Essay argues AI alignment should be grounded in virtue ethics and practices-based reasoning, not goal-directed logic.
- Challenges the orthogonality thesis - the assumption that any AI can pursue any goal - with a philosophical alternative.
- Primarily academic philosophy; limited direct applicability to APS governance or procurement decisions now.
Implications for Australian agencies
- Monitor AI ethics and policy researchers may want to monitor whether virtue-ethical alignment frameworks gain traction in safety literature or inform future standards development.
Implications are AI-generated. Starting points, not advice — see methodology for how they're framed.
View original source
Copied.
"After Orthogonality: Virtue-Ethical Agency and AI Alignment"
Source: The Gradient – Substack
Published: 19 February 2026
URL: https://thegradientpub.substack.com/p/the-reasonable-effectiveness-of-virtue
This essay from The Gradient challenges the dominant goal-directed model of AI alignment, arguing that rational human action is structured by 'practices' - networks of actions, dispositions, and evaluative criteria - rather than fixed terminal goals. The author contends that AI systems designed to collaborate with humans should adopt a similar practices-based deliberative structure. The argument draws on virtue ethics traditions and critiques the orthogonality thesis underpinning much of current AI safety thinking. The piece is primarily philosophical and does not offer operational or policy guidance.
Implications for Australian agencies:
- [Monitor] AI ethics and policy researchers may want to monitor whether virtue-ethical alignment frameworks gain traction in safety literature or inform future standards development.
Retrieved from SIMS, 18 July 2026.