The COVID-19 pandemic forced a sudden shift to fully online delivery of learning in many higher education institutions across the world in 2020. Many academics used pre-recorded videos as a replacement for live in-person lectures. It can be time consuming to create these lecture videos and to update them for subsequent course offerings. In this practice paper, we present an approach which automates the creation of lecture videos using presentation slides with speaker notes text. We implement an Automated Video Presenter (AVP) that uses images of the presentation slides as the visual element of the video and uses text-to-speech (TTS) technology to synthesise speech from the speaker note text. Having used this approach and the AVP-generated videos for two semesters, we discuss the benefits and the issues associated with them, as well as the lessons learnt. More importantly, we see wide ranging implications for both pedagogy and research and future work.