COMP90087 · The Ethics of Artificial Intelligence
Virtue Ethics
Week 5 completes the three-theory toolkit with character-based ethics: Aristotle’s doctrine of the mean, phronesis (practical wisdom), plus Shannon Vallor’s technomoral virtues and care ethics (Gilligan, Tronto). The exam rewards two things here — matching a virtue to its deficiency-vice and excess-vice, and cleanly attributing concepts (phronesis = Aristotle; technomoral virtues = Vallor; Gilligan/Tronto = care ethics). Virtue ethics is also the lens marked answers often favour for high-stakes automation cases.
What this chapter covers
- 01Three-theory contrast: utilitarianism = consequences, deontology = duty, virtue ethics = character
- 02Ethical (character) vs intellectual virtue; virtues as settled dispositions acquired by habit and modelling
- 03Eudaimonia (flourishing) and Hursthouse’s criterion: an action is right if it is what a virtuous agent would characteristically do
- 04The doctrine of the golden mean — virtue as the context-relative mean between a vice of deficiency and a vice of excess
- 05Phronesis (practical wisdom) — the master virtue that applies and harmonises the others in concrete situations
- 06Shannon Vallor’s 12 technomoral virtues (incl. technomoral wisdom = phronesis) and techno-social opacity
- 07Care ethics (Gilligan, Tronto): care as a virtue, in the right amount and to the right people; a reaction against purely impartial rule-based ethics
- 08Attribution traps: phronesis/golden mean = Aristotle; technomoral virtues / “the AI Mirror” = Vallor; Gilligan & Tronto = care ethicists
Locate the virtue as the mean between two vices, then attribute correctly
- +1Courage. Deficiency = cowardice (too little confidence/anger in the face of danger); virtue = courage; excess = rashness. The mean is context-relative — the “right” amount of courage differs for a minor risk versus a grave one.
- +1Truthfulness. Deficiency = self-deprecation (understating yourself); virtue = truthfulness about oneself; excess = boastfulness. Note the deficiency here is not “lying” but under-claiming.
- +1Generosity (liberality). Deficiency = stinginess/penny-pinching; virtue = generosity; excess = extravagance/being a spendthrift. Getting the deficiency-vs-excess side correct is what earns the mark in the matching item.
- +1Attribution. Phronesis (practical wisdom) and the doctrine of the mean are Aristotle’s; the 12 technomoral virtues and “the AI Mirror” are Shannon Vallor’s (her technomoral wisdom = Aristotle’s phronesis, updated for technology). Gilligan and Tronto belong to care ethics, not to Aristotle or Bentham.
Key terms
- Virtue ethics
- Character-based ethics: it asks “what sort of person should I try to be?” rather than only which act or outcome or duty is right. Rooted in Aristotle (and Confucius, Buddha).
- Doctrine of the golden mean
- Aristotle’s idea that a virtue is the context-relative mean between a vice of deficiency and a vice of excess (e.g. courage between cowardice and rashness). It is not a literal arithmetic midpoint.
- Phronesis (practical wisdom)
- Aristotle’s master virtue, needed to apply the other virtues to concrete situations and to harmonise them when they conflict. Vallor recasts it as “technomoral wisdom” for technology.
- Eudaimonia
- Flourishing — a fully good and fortunate human life, of which the virtues are constitutive. A bad person may feel pleasure but cannot truly be eudaimon.
- Technomoral virtues (Vallor)
- Shannon Vallor’s 12 virtues for a high-tech age (honesty, self-control, humility, justice, courage, empathy, care, civility, flexibility, perspective, magnanimity and technomoral wisdom), a response to growing techno-social opacity.
- Care ethics
- A feminist virtue tradition (Gilligan, Tronto) treating care — attending, listening, taking responsibility — as a virtue enacted in the right amount and to the right people, and reacting against purely impartial, rule-based ethics.
Virtue Ethics FAQ
How do I not mix up which side of the mean a vice is on?
Fix the pattern deficiency → virtue → excess and test every pair against it: cowardice–courage–rashness, stinginess–generosity–extravagance, self-deprecation–truthfulness–boastfulness. The matching MCQ specifically checks whether you can say a vice is the deficiency or the excess, so practise a handful until the pattern is automatic. Remember the mean is context-relative, not the arithmetic middle.
What’s the exam trap between Aristotle and Shannon Vallor?
The exam tests attribution: phronesis, the golden mean and eudaimonia are Aristotle’s; the 12 technomoral virtues and “the AI Mirror” are Vallor’s (her ‘technomoral wisdom’ is phronesis updated for technology). Care ethics belongs to Gilligan and Tronto — not Bentham or Kant. Keep a who-owns-which-term card for these.
Why do marked answers often prefer virtue ethics for automation cases?
Because virtue ethics asks what a person of practical wisdom would do and which character traits a choice expresses, it handles “should we let AI make this high-stakes call?” well — you can weigh justice, care and over-trust in automation together and invoke phronesis to adjudicate. It is flexible and situation-sensitive, though its critics call it vague; the reply is that the golden mean plus phronesis give it direction.
Can AI help me revise virtue ethics for COMP90087?
Yes. Sia can drill the deficiency–virtue–excess table, quiz you on Aristotle-vs-Vallor-vs-care-ethics attribution, and coach a virtue-ethics run on a fresh automation case — explaining each step and checking your reasoning. It mirrors how the University of Melbourne assesses the material and does not produce work you submit; the subject’s GenAI and integrity rules apply.
Exam move
Build two artefacts: a golden-mean table (deficiency → virtue → excess for a dozen traits) and a strict attribution card (Aristotle = phronesis / mean / eudaimonia; Vallor = technomoral virtues / AI Mirror; Gilligan & Tronto = care ethics). The exam almost always tests both a matching item and a “who said it” item, so make these pure recall. For essays, rehearse a virtue-ethics run — pick two or three relevant virtues, ask what a person of phronesis would do, and weigh the character the choice expresses — since this is the lens marked answers favour for high-stakes automation. Fold it into weekly revision so it is solid before the closed-book hurdle exam.
Working through Virtue Ethics in COMP90087? Sia is AskSia’s AI AI Ethics tutor — ask any COMP90087 Virtue Ethics question and get a clear, step-by-step explanation grounded in how COMP90087 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
