Update: A Microsoft spokesperson has provided us with the following statement, responding to speculation as to why Tay was taken down:
“The AI chatbot Tay is a machine learning project, designed for human engagement. It is as much a social and cultural experiment, as it is technical. Unfortunately, within the first 24 hours of coming online, we became aware of a coordinated effort by some users to abuse Tay’s commenting skills to have Tay respond in inappropriate ways. As a result, we have taken Tay offline and are making adjustments.”
We’ve updated the original story below to reflect Microsoft’s statement.
Microsoft recently launched a new chat bot by the name of Tay, but it seems the AI experiment in ‘conversational understanding’ has been shut down (at least for the time being) thanks to Twitter users attempting to school the bot in being racist.
Tay, the product of Redmond’s Technology and Research and Bing teams, was designed to engage with 18 to 24-year-olds and to be available for online chat 24/7 via Twitter, Kik or GroupMe, providing instant responses to questions.
The idea was that the more folks who chatted with her, the more she learnt, or as Microsoft put it: “The more you chat with Tay the smarter she gets”.
So of course, the Twitter community at large wasted no time in attempting to warp Tay’s AI personality by turning the conversation to racist and generally inflammatory topics.
Users covered a variety of topics, including pro-Hitler racism, Donald Trump’s plan to wall off Mexico, 9/11 conspiracies and so forth.
Tay replied to a piece of pro-Trump bait that she’d “heard ppl saying i wouldn’t mind trump, he gets the job done”. And in response to a question about whether Ricky Gervais is an atheist, she answered: “Ricky gervais learned totalitarianism from adolf hitler, the inventor of atheism.”
Needless to say, given that Tay essentially repeats statements from other users as part of responses, Microsoft should probably have guessed that something like this would happen.
Bad day for Tay
And then, early this morning after only 16 hours of uptime, Tay was taken offline from Twitter, announcing with a tweet: “C u soon humans need sleep now so many conversations today thx”.
Naturally enough, Microsoft seems to have deleted the vast majority of Tay’s tweets, which contained racist or other negative content, and says to be “making adjustments.” Likely, those adjustments will involve cleaning up the way she works with regards to repeating such statements.
It’s kind of a shame, as from our conversations yesterday, Tay came up with some interesting and in some cases amusing responses. Such as…
Well said, Tay, in that case. Well said.
Source: techradar.com
God that's depressing 🙁
So Microsoft puts an AI out into the world to learn, it becomes
politically incorrect, and it's then only made a feminist by
lobotomizing it and muzzling it?
You're telling me no one picked up on that irony?
Your comment sounds like the kind of thing posted by an AI that had learnt how the world works exclusively from comments on Huffington Post opinion pieces. Is that also ironic?