Answers are grounded in this paper and cite their sources.
Forecasts about AI and education tend to swing between promise and alarm, and most of them will be wrong. Scenario planning offers a more useful discipline: building several plausible futures, testing today's plans against each of them, and identifying the choices that decide which future arrives. This article draws on two scenario series written by the first author for educators in 2026, one looking to 2035 and one to 2045, together with notes from UNESCO's Education Beyond 2030 consulta‐ tion. It reads these alongside published scenario sets from the UK Gov‐ ernment Office for Science, the World Economic Forum, the OECD, UN‐ ESCO and EdTech Hub. It sets out seven pressures already shaping educa‐ tion, each labelled by the strength of its evidence; four futures for 2035 built on two uncertainties, the pace of AI and whom it serves; and three futures for 2045, from a "great unbundling" through a two-tier system to a "great release". It identifies six switches, decisions being made now about learner records, funding for human teachers, assessment, owner‐ ship of models and data, participation and accountability, which more than AI capability will determine who benefits. It closes with moves that hold up in every future, signals to watch, and a backcast of decisions needed before 2035.
No takes yet. Share an insight, caveat, or question.
paul alan penfold (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: