We began this series with Course 1’s invitation to think before we build. We have spent eleven courses developing the analytical tools to think more deeply, more honestly, and more rigorously than the dominant frameworks of AI ethics typically allow.
This final lesson asks: what does it mean to make critical practice ongoing, not a one-time encounter with theory, but a sustained orientation toward AI ethics that continues to develop as circumstances change?
The three pillars of Ethos Sophia revisited.
Ethos, character, values, the ethics that guide choices. The critical traditions in this series have expanded our understanding of what ethical character requires in the context of AI. It requires not just following rules, but developing the habit of asking whose interests are served, whose are ignored, and what structural forces shape the options available. Ethical character in AI is not comfortable. It requires raising concerns that may be unwelcome, questioning assumptions that may be convenient, and advocating for people whose voices are absent from decision-making.
Sophia, wisdom earned through reflection, not just knowledge. Knowledge of CRT, feminist theory, Marxist theory, and postmodern theory is necessary but insufficient. Wisdom is what we develop through the repeated practice of applying these frameworks to real situations, through the discipline of noticing what we had not previously noticed, of asking questions we had not previously thought to ask, of sitting with complexity rather than reaching for premature resolution.
Praxis, putting wisdom into action in the real world. This is the most demanding pillar, because it is the one that involves risk. Asking critical questions in organisational contexts is not cost-free. Advocating for affected communities whose interests diverge from those of your employer is not always easy. Challenging algorithmic authority when it produces unjust outcomes requires courage as well as competence.
The ongoing character of critical practice. AI systems are not static. They are continuously updated, retrained, and redeployed. The social contexts in which they operate change. The regulatory landscape evolves. New AI applications raise ethical challenges that existing frameworks were not designed to address. Critical practice must be ongoing to remain relevant, not a set of principles applied once and then forgotten, but a discipline of continuous attention and renewal.
The community dimension. One mind cannot see it all, and one critical tradition cannot see it all either. The most powerful form of critical AI practice is collective: communities of practice in which people with different disciplinary backgrounds, different professional experiences, and different social positions bring their critical tools to shared problems. Ethos Sophia is an attempt to build such a community, not a community of agreement, but a community of rigorous, honest, mutually accountable inquiry.
A note on humility. The critical traditions in this series have their own blind spots, their own limitations, their own exclusions. CRT has been critiqued for centring American racial dynamics in ways that travel poorly to other contexts. Feminist theory has historically centred the experiences of white Western women. Marxist theory has been used to justify forms of oppression as severe as those it critiques. Postmodern theory can become a sophisticated form of inaction.
The appropriate response to these limitations is not to abandon the frameworks but to use them with humility, attentively to their limits as well as their insights, in combination rather than isolation, and always in dialogue with the people most affected by the systems we are analysing.
Final reflection question: What one commitment to critical AI practice will you carry from this series into your work? Not an abstract principle, but a concrete, specific commitment, something you will do differently because of what you have encountered here.