COMP90087 · The Ethics of Artificial Intelligence
Utilitarianism
Week 3 introduces the first of the three moral theories: consequentialism and the principle of utility — the right action maximises overall happiness (pleasure over suffering) counted impartially across everyone affected. This is the theory behind the exam’s one genuinely quantitative item, where you compute net utility across options and pick the maximum, and it is a standard lens for both essays. Master the vocabulary (act vs rule, welfarism, aggregation) and the standard objections so you can apply it fast and defend it.
What this chapter covers
- 01Utilitarianism = consequentialism: only outcomes matter, measured in happiness/wellbeing (Bentham, Mill, Singer)
- 02The principle of utility — maximise the greatest total balance of happiness over unhappiness across all affected, counted impartially
- 03Core features: consequentialism, welfarism/hedonism, impartiality and aggregation
- 04Act utilitarianism (evaluate each act) vs rule utilitarianism (follow the rules whose general adoption maximises utility)
- 05The application procedure for AI cases: identify affected parties → estimate happiness/suffering per option → sum → pick the greatest net utility
- 06Weighing consequences by magnitude × probability × scope (avoiding ad-hoc weighing)
- 07Standard objections: can justify violating individual rights / using people as mere means; demandingness; measurement and prediction problems; ignores distribution and justice
Apply the principle of utility to a feature-rollout choice
- +1State the method. Net utility = (beneficiaries × gain) − (annoyed × loss) for each option; the utilitarian ships the option with the greatest net total. Everyone’s satisfaction counts equally (impartial aggregation).
- +1Feature A: (300 × +2) − (200 × 4) = 600 − 800 = −200. A net-negative option — it should not ship despite pleasing 300 people.
- +1Feature B: (150 × +6) − 0 = +900. Feature C: (500 × +3) − (400 × 3) = 1500 − 1200 = +300.
- +1Compare and select the maximum: B (+900) > C (+300) > A (−200). Ship Feature B. The teaching point: breadth of benefit does not settle it — C reaches the most users (500) but its large annoyance cost sinks it; net balance decides.
Key terms
- Principle of utility
- The utilitarian criterion (Bentham/Mill): the right action produces the greatest total balance of happiness over unhappiness across everyone affected, counted impartially. Morality is the happiness of beings in this world, nothing more.
- Consequentialism
- The view that only consequences — the resulting states of affairs — determine rightness, in contrast to deontology’s focus on duties and virtue ethics’ focus on character.
- Welfarism / hedonism
- The value theory behind utilitarianism: what ultimately matters is happiness or wellbeing (pleasure over suffering), so utility is measured in those terms.
- Impartiality and aggregation
- Everyone’s happiness counts equally and is summed — the “greatest good for the greatest number.” This is also the source of the theory’s main objection: minorities’ interests can be outweighed.
- Act vs rule utilitarianism
- Act utilitarianism evaluates each individual act by its consequences; rule utilitarianism follows the rules whose general adoption would maximise utility. Naming the variant is the first move in an essay.
- Demandingness objection
- The criticism that utilitarianism can require enormous personal sacrifice (always maximise the good) and can license violating individual rights if the aggregate payoff is large enough.
Utilitarianism FAQ
How do I actually compute utility in the exam’s numeric question?
For each option compute net utility = (number pleased × per-person gain) − (number annoyed × per-person loss), then pick the option with the highest total. The classic trap is to choose the option that helps the most people rather than the one with the greatest net balance — always work out every option’s number before selecting.
What’s the difference between act and rule utilitarianism?
Act utilitarianism judges each individual action by whether it maximises utility; rule utilitarianism asks which general rules, if everyone followed them, would maximise utility, then follows those rules. Rule utilitarianism can avoid some act-utilitarian counter-examples (e.g. it can forbid framing an innocent person as a standing rule). Naming which variant you are using is a marked essay move.
What are the standard objections I should be able to raise?
That it can justify violating individual rights or using people merely as means if the aggregate benefit is high; that it is over-demanding; that measuring and predicting happiness is hard; and that it ignores distribution and justice (a huge benefit to a majority can be counted as outweighing serious harm to a minority). These objections are exactly where deontology and virtue ethics push back in Weeks 4–5.
Can AI help me practise utilitarian reasoning for COMP90087?
Yes. Sia can set you fresh net-utility problems with new numbers, walk through the magnitude × probability × scope weighing for an essay, and check that your verdict follows from the sums — explaining each step. It is a study aid built to mirror how the University of Melbourne assesses this, not a substitute for your own graded work, and the subject’s academic-integrity and GenAI rules apply.
Exam move
Drill the net-utility computation until it is automatic — for every option write (pleased × gain) − (annoyed × loss), then choose the maximum, and rehearse the trap that breadth of benefit does not win. For essay use, practise the qualitative version: weigh harms and benefits by magnitude, probability and scope rather than ad-hoc. Keep the vocabulary crisp (consequentialism, welfarism, impartial aggregation, act vs rule) because attribution and definition MCQs reward it, and pair the theory with two standard objections so you can pre-empt the rebuttal your essay needs. This is examinable both as the numeric MCQ and as an essay lens, so build both the calculation and the argument.
Working through Utilitarianism in COMP90087? Sia is AskSia’s AI AI Ethics tutor — ask any COMP90087 Utilitarianism 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.
