The Turing Podcast: how data science and machine learning are answering questions about human evolution

Wednesday 21 Jul 2021

In the latest episode of The Turing Podcast, our hosts Bea Costa Gomes and Ed Chalstrey are joined by Robert Foley, Professor of Human Evolution at the University of Cambridge and a Turing Fellow. The episode focuses on his  ‘Palaeoanalytics’ Turing project, and considers how our understanding of human evolution has changed in recent decades.

 Analysing the stone tools used by ancient humans can provide a multitude of insights into how our ancestors lived and interacted with their environment. However, this analysis is a highly skilled and time-consuming task – it can take over 15 minutes per tool, and some samples consist of hundreds of thousands of tools.

The Palaeoanalytics project is developing AI and machine learning approaches to ‘palaeo-data’ (data from the deep past), which will provide new ways of automatically recognising and categorising the sizes and shapes of stone tools, as well as other prehistoric artefacts such as the fossils of extinct animal and human species.

Podcast guest Robert Foley said: 

“There is a vast amount of evidence, stone tools and of publications on the topic and I wanted to find a way to bring that together. Analysing stone tools is interesting but quite a slow process – what can we do to automate this and up the game?” 

Listen to the podcast here, or subscribe via Spotify, ApplePodcasts, Stitcher, Podbay, Podbean, iHeartRadio or Listen Notes