Schools are being asked to prepare young people for economies, technologies and civic conditions that nobody can describe with confidence. Estimates of how automation and generative artificial intelligence will alter work differ widely, yet most of them point in one direction: tasks built on routine procedure are the easiest to hand over to machines, while tasks that call for judgement, explanation and cooperation with others are gaining value. This conceptual paper asks what kind of education answers that situation, and it argues that the long tradition of liberal education offers a better starting point than the narrow, examination-driven specialisation that still governs much schooling in India and elsewhere. The paper first reviews labour-market and policy evidence on the changing demand for skills, drawing on Autor, Levy and Murnane, on Deming, and on Eloundou and colleagues, along with the framing documents of UNESCO and the OECD. It then traces the idea of liberal education from the classical artes liberales, through Newman and Dewey, to Nussbaum's account of democratic citizenship, and sets this history alongside the Indian precedents recognised in the National Education Policy 2020. A review of empirical studies follows. Research from the Wabash National Study and related work links particular liberal arts experiences, such as clear teaching, sustained discussion and frequent contact with teachers, to gains in critical thinking, while meta-analytic evidence shows that critical thinking responds to explicit instruction set in authentic problems. The same literature cautions that thinking skills do not transfer automatically across domains and that much of the evidence rests on self-selected samples and self-report. From this evidence the paper develops a distinction between structural integration, in which one discipline supplies the working logic of a task in another, and decorative integration, in which subjects are simply placed side by side around a theme. It argues that liberal education earns its claims only in the first form. The paper closes with implications for Indian school education and teacher preparation under NEP 2020 and the National Curriculum Framework for School Education 2023, and it identifies examination design and teacher education as the places where reform is most likely to stall.