When a machine makes a decision that harms someone, who is answerable — and how would you even decide what ‘good’ means?
Case 6 is the other side of the whole module. Every advance that finds a cure also displaces work, bakes in bias, and erodes privacy — and self-driving cars sharpen it to a single question: good for whom?
The same innovation that finds a cure also brings a cost, and it is dishonest to skip it: job displacement; work stripped of what made it human; knowledge hoarded or hidden; inequality as gains pool at the top; bias baked quietly into a model; privacy worn away a little at a time; and the hard question underneath — when a machine makes the call, who is answerable when it gets it wrong?
The exam-ready list of harms: displacement, dehumanised work, knowledge hoarding, inequality, bias, privacy erosion, and the accountability gap.
When a self-driving car causes harm, who is answerable — the person in the seat who was not driving, the company, or the algorithm? The courts are still working it out. This is where Khogali & Mekid (2023) plant their flag: they weigh it the utilitarian way (greatest good for the greatest number) and through social-impact theory — a formal way of asking the question the whole module circles: good for whom?
The real question was never efficiency or people, as if you had to pick a side. On one side: cures found faster, discovery at superhuman scale. On the other: displacement, and a widening gap in who holds power. The answer is not to choose a side but to keep the whole scale balanced on people — the through-line of every framework in the module.
Industry 5.0, Society 5.0, human-centred design — each in its own language says the same thing: the innovation that lasts is the kind pointed at people. The shape was always there; we just had to notice it.
List the other-side costs of AI-driven innovation.
What do the self-driving figures show (AAA, 2025)?
How do Khogali & Mekid (2023) weigh the question?