Projet Oxford Microsoft

Microsoft stops research on emotion-reading AIs

Definitely, conversational or interactional AIs are no longer popular among American tech giants. Google still refuses to release its chatbots the most advanced because of the risk of sexist and racist biases, Facebook dropped its visual identification program, not to mention the huge scandal Clearview AI. Microsoft has in turn announced that it is abandoning the development of the oxford projectits facial recognition AI capable of “reading” human emotions.

Microsoft Oxford Project

AI had been in development since 2015 and Microsoft had even made it accessible through its service Azure. The Redmond firm now believes that the new ethical rules put in place by the company no longer allow AI to be kept in operation. Once again, AI is singled out for its lack of inclusiveness: the algorithm of the Oxford project would be unable to adapt its reading of facial emotions to different human cultures.

In addition, creating an emotion-reading AI would require collecting a huge stock of data… a constraint seen with a rather dim view after the Cambridge Analytica and Clearview AI scandals. Finally, Microsoft’s AI classified only men and women by gender, a lack of “gender-fluidity” which obviously poses a problem (and this even if 98% of the human population defines its gender according to its biological sex ). Scraps of the Oxford Project will continue to exist in Seeing AIa visual recognition technology that helps describe a scene for the visually impaired.

Leave a Comment

Your email address will not be published.