Chapter 13. Work, Skills, Leadership, and the Future of AI

13.1 Tasks change before whole occupations

AI usually affects tasks within jobs rather than replacing every responsibility in an occupation at once. Some tasks may be automated, others accelerated, and new tasks created around checking, coordinating, designing, and governing systems. The effect varies by sector, workflow, regulation, customer expectations, and the availability of complementary skills.

Workers and employers should analyse tasks rather than assume that a job title predicts the outcome. A useful task inventory identifies routine information processing, judgement-intensive work, interpersonal responsibilities, physical tasks, and activities requiring accountability.

13.2 Skills for an AI-enabled workplace

Important skills include domain expertise, critical thinking, communication, data literacy, problem formulation, digital security, ethical reasoning, collaboration, and the ability to evaluate AI output. Technical roles may require programming, statistics, model evaluation, data engineering, or AI operations. Leadership roles need strategy, governance, procurement, and change management.

People should practise asking good questions, checking evidence, recognising when a result is outside their expertise, and explaining how they used AI. These capabilities are transferable across providers and model versions.

13.3 Leadership and organisational culture

Leaders should communicate why AI is being introduced, what it will and will not do, how staff will be trained, and how concerns will be addressed. Involving employees early can uncover workflow realities, privacy concerns, and practical failure modes that senior decision-makers may miss.

AI should not be introduced as a reason to eliminate human consultation or impose opaque monitoring. Workforce changes should be discussed fairly, with attention to retraining, job quality, accessibility, and the distribution of productivity gains.

13.4 Human judgement and professional identity

AI may provide a useful second perspective, but professional judgement includes responsibility for context, values, relationships, and consequences. In education, a teacher supports learning and development; in healthcare, a clinician interprets evidence in a patient context; in management, leaders are accountable for choices and their effects.

The goal is not to pretend that AI has no value or that it can do everything. The goal is to allocate tasks sensibly and preserve the human capabilities that matter.

13.5 Future scenarios

Future AI systems may become more capable, integrated, multimodal, and agentic. Some capabilities may improve quickly, while others remain unreliable or expensive. Predictions about timelines and economic effects are uncertain, so planning should use scenarios rather than single-point forecasts.

Organisations can prepare by improving data governance, building adaptable skills, designing modular systems, monitoring standards and regulation, and avoiding unnecessary lock-in. The best strategy is to remain ready to learn while making decisions based on evidence.