Evaluate the benefits and risks of artificial intelligence and automation for work, society and human agency, with balanced arguments and examples

What this dot point is asking

This theme gives you the balanced knowledge to argue any General Paper question on artificial intelligence and automation: their effect on jobs, the economy, fairness and human agency. The central insight is that AI is neither inherently a saviour nor a threat; its net effect on society depends on how it is designed, governed and distributed. A strong answer marshals the benefits and the risks, then judges that the outcome turns on human choices, not on the technology alone.

The answer

The case for AI and automation

The benefits are substantial and worth developing precisely:

• Productivity and growth. Automating routine work raises output and can lift living standards across an economy.
• Augmentation, not just replacement. AI extends human capability, supporting medical diagnosis, accelerating scientific research and handling scale no human team could.
• Access and convenience. AI-driven services deliver translation, tutoring and information to people who could not otherwise afford them.
• Relief from drudgery. Offloading repetitive tasks can free people for more creative, interpersonal or judgement-based work.

The case for caution

The risks are equally real:

• Displacement and inequality. Automation now reaches routine cognitive as well as manual jobs, and its gains may flow to those who own the technology while costs fall on displaced workers.
• Bias and opacity. Algorithms trained on biased data can entrench discrimination, and their "black box" nature makes decisions hard to challenge.
• Concentration of power. AI capability clusters in a few large firms and states, raising concerns about surveillance and accountability.
• Erosion of human skill and agency. Over-reliance can dull judgement, from navigation to clinical decisions, and outsource choices we should make ourselves.

Augmentation versus replacement

A crucial distinction for nuanced answers: most AI deployment so far augments human work rather than wholly replacing it. A radiologist using AI to flag scans is augmented; a fully automated checkout replaces a cashier. Recognising that the future of work is a mix, not a single story of mass unemployment, lets you avoid both techno-utopian and apocalyptic extremes.

The outcome depends on governance

The decisive move in a GP essay is to argue that AI's effect is not determined by the technology but by human choices: regulation of bias and safety, investment in reskilling, and how the gains are distributed. This is why the same technology can widen inequality in one society and broadly raise welfare in another. It also lets you attack absolutes like "always" or "inevitably" in a question.

Examples in context

Example 1. Singapore's national AI strategy. Singapore has pursued a national strategy to deploy AI across sectors such as healthcare, transport and government services, while emphasising trusted and responsible use and heavy investment in workforce reskilling through initiatives like SkillsFuture. This illustrates the governance argument in action: the state treats AI's benefits as conditional on managing its risks and on equipping workers to adapt, a concrete counter to any claim that technology's effects are simply given.

Example 2. Algorithmic bias in decision-making. Documented cases worldwide of AI systems producing biased outcomes in hiring, lending or policing, because they learned from skewed historical data, show why opacity and bias are not abstract worries. They evidence the risk side of any AI essay and support the argument that benefits depend on regulation: without scrutiny and accountability, an efficient system can scale unfairness as readily as it scales convenience.

Try this

Q1. Explain the difference between AI augmenting and replacing human work, with an example of each. [2 marks]

• Cue. Augmentation extends a worker's capability (a doctor using AI to flag scans); replacement removes the role entirely (an automated checkout replacing a cashier).

Q2. Identify one reason AI's economic gains might worsen inequality. [2 marks]

• Cue. The gains may flow to those who own the technology and capital, while displaced workers bear the costs, so without redistribution or reskilling the gap between them widens.

Q3. Explain why "the benefits of AI will always outweigh its dangers" is vulnerable as a claim. [3 marks]

• Cue. "Always" is an absolute that ignores cases where unmanaged AI causes net harm and assumes benefits are evenly shared; since the balance depends on governance and distribution, the outcome is conditional, not guaranteed.