A partnership between the British Library and Alan Turing Institute with data scientists, curators, research software engineers, computational linguists, digital humanities scholars and historians from those institutions and universities including Exeter, University of East Anglia, Cambridge and Queen Mary University of London, the Living with Machines project is developing data science and AI methods to ask historical questions using digitised collections at scale. Our sources include millions of pages of historical newspapers, novels, maps, census records, directories and other sources. We hope that the research methodologies and tools developed through the project will be adapted and used by cultural heritage professionals and researchers to access and understand digitised historic collections in the future. This talk will outline why the British Library sought to collaborate with the UK's data science and AI institute. It will share some early methodological results from the project, and reflect on lessons learnt from working with an interdisciplinary team to apply data science methods for research questions in areas as varied as computational linguistics, human computing/crowdsourcing, historical analyses of space and time, data science and software engineering. It will also consider the implications of computational metadata generation or enhancement for existing cataloguing and discovery systems within the Library, and discuss our efforts to share work in progress with staff across the British Library.