It took just one weekend for Meta's new AI Chatbot to become racist

mashable.com

17 points·Grimburger··7 comments

(Circa : 2016)

search: >'MicroSoft AI Chatbot Tay became racist, sexist'< @DDG : <https://html.duckduckgo.com/html/?q='MicroSoft AI Chatbot Ta...>

When the racism is everywhere you look; no matter what you do to remove it... You might need to look at yourself.

Kindof like "if everyone you know is an asshole; its likely none of them actually are"

It is one of those things. There is the idea that one day we will make an AI that is far superior to us, and yet current AI seems to be hopelessly reflective of biological being that create them.

Some something make something more than it self, like a gravity assist? One of the big questions.

In order for an AI to understand the world like a human, it will need to interact with the world. Absent the ability to perform real world experiments, there can be no closure to the hypothesis => experiment => conclusion loop that is needed to build up an accurate model of reality.

I wonder how often it actually says racist things and what kind of context it says them in. I just tried it out and mostly it just went off on its own topic. I kept having to redirect it. But it didn’t say anything racist to me.

"Ïnline training model assimilates inline training data", more news at 11!

It took just one weekend for Meta's new AI Chatbot to accurately mimic FB userbase.

Reply on news.ycombinator.com