Within 2 months, Google fired its two co-directors of AI ethics, Timnit Gebru and Margarett Mitchell. During this time, Google has also developed, and will probably soon deploy, algorithms trained on huge amounts of unsecured data, in an extremely opaque way, which had been criticized by the two co-directors in a research paper.
The hasty deployment of very dangerous algorithms without any internal or external audit, on the phones of billions of humans, seems to me absolutely terrifying. But the scariest part is that this story is completely ignored by the general public, the media, the tech industry and the academic world... #AISafety #AIEthics #LeFabuleuxChantier
Script + sources : https://docs.google.com/document/d/e/2PACX-1vSo9ZvnYWMitxXb79ZGLeCukOJ7OIHby5TCwH7kQw5JBER_92P-G8g2yZk_HkTWjfMURfutLBf6UPnU/pub
A thread on the topic: https://twitter.com/le_science4all/status/1334487647033696256
Twitter : https://twitter.com/le_science4all
Facebook : https://www.facebook.com/Science4Allorg/
La formule du savoir (mon 1er livre) :
https://laboutique.edpsciences.fr/produit/1035/9782759822614/La%20formule%20du%20savoir
Le fabuleux chantier (mon 2e livre, avec El Mahdi El Mhamdi) :
https://laboutique.edpsciences.fr/produit/1107/9782759824304/Le%20fabuleux%20chantier
Turing à la plage (mon 3e livre, avec Rachid Guerraoui) :
https://www.dunod.com/sciences-techniques/turing-plage-intelligence-artificielle-dans-un-transat
Probablement? en audio : http://playlists.podmytube.com/UC0NCbj8CxzeCGIF6sODJ-7A/PLtzmb84AoqRQ0ikLb4yC4lKgjeDEIpE1i.xml
Moi en podcast avec Mr Phi :
Version YouTube : https://www.youtube.com/channel/UCNHFiyWgsnaSOsMtSoV_Q1A
Version Audio : http://feeds.feedburner.com/Axiome
Sous-titres sur les autres vidéos : http://www.youtube.com/timedtext_cs_panel?tab=2&c=UC0NCbj8CxzeCGIF6sODJ-7A