A massive effort by African volunteers is ensuring artificial intelligence understands their native languages, and measuring 40,000 skeletons

Our AI summer continues with a look at how to get artificial intelligence to understand and translate the thousands of languages that don’t have large online sources of text and audio. Freelance journalist Sandeep Ravindran joins host Sarah Crespi to discuss Masakhane, a volunteer-based project dedicated to spurring growth in machine learning of African languages. See the whole special issue on AI here.

Also this week on the show, Eucharist Kun, a Ph.D. student at the University of Texas at Austin, and colleagues used machine learning to take skeletal measurements from x-rays stored in the UK Biobank. Kun discusses links from these body proportions to genes, evolution, and disease.

Finally, in a sponsored segment from the Science Custom Publishing Office, Erika Berg, director and senior editor of custom publishing, interviews Aysha Akhtar, co-founder and CEO of the Center for Contemporary Sciences, about how the Food and Drug Administration Modernization Act 2.0 along with advances in technology are clearing the way for alternatives to animal testing in the development of new drugs. This segment is sponsored by Michelson Philanthropies.

This week’s episode was produced with help from Podigy.

About the Science Podcast

Authors: Sarah Crespi; Sandeep Ravindran

Episode page: https://www.science.org/doi/10.1126/science.adj7646

Learn more about your ad choices. Visit megaphone.fm/adchoices

Adding thousands of languages to the AI lexicon, and the genes behind our bones 38:59 Share

Science Magazine Podcast

Shownotes Transcript

Adding thousands of languages to the AI lexicon, and the genes behind our bones