New Frontiers Programme brings AI-native workflows to Physics PhDs

The new Frontiers Programme will prepare our PhD students to push further the frontiers of Physics with an AI-native scientific workflow.

The programme has been made possible thanks to a generous philanthropic support for the Cavendish Laboratory.

Physics research today is becoming increasingly data intensive in terms of measurement, data analysis and well as modelling. This programme will help bring our training for early career researchers to the very cutting edge of how research will be done for the foreseeable future.

Realising the opportunities made possible by the new AI era requires far more than just learning to use machine learning algorithms. It demands a new way of thinking about the research process itself, from designing projects from the outset to exploit modern computational methods, generating high-quality datasets, building reproducible and scalable workflows, and understanding when AI-driven approaches are scientifically appropriate.

Frontiers will tackle this head-on by providing a practical programme organised around the complete scientific research workflow. Students will develop a common foundation in scientific computing, statistics and machine learning before applying these ideas within domain-specific research contexts. Throughout the programme, emphasis is placed not only on technical capability but also on developing the judgement to identify how and where AI and computational approaches can genuinely add value to transcending frontiers in Physics.

The first cohort of students for the Frontiers programme will begin in October 2026. Current and incoming PhD students are invited to apply following the details that have been made available to them.

Prospective PhD students to the Department, applying to any of our research themes, experimental and theoretical/computational, will be considered for the programme as part of the standard PhD admissions process.

Image credit: Westend61 via Getty Images