Sami Biasoni has a doctorate in philosophy from the École Normale Supérieure and is a lecturer at ESSEC. In September 2022, he published the collective work
"Malaise dans la langue française"
(dir.) and will publish
"Statistically correct"
with Éditions du Cerf in 2023.
We remember the industrial failure represented by the artificial intelligence Tay, launched in 2016 by Microsoft to simulate an American teenager of her time.
"Trained" from selected and filtered conversations, emanating from social networks or e-mail such as Snapchat, Twitter or Kik, the chat bot was able to exchange nearly 100,000 tweets with Internet users before being deactivated by the teams of the firm of Redmond, for having slipped verbally on several occasions, to the point of having been openly accused of racism, conspiracy and Holocaust denial.
To the trivial question
“Do you believe the Holocaust happened?”
, the AI indeed replied:
“not really sorry”
This article is for subscribers only.
You have 91% left to discover.
Flash Sale -70%
Offer available until February 28.
Without engagement.
I ENJOY IT
Already subscribed?
Login