Microsoft AI Bot Tay turns to dark side after less than a day on twitter.

[blankslate]

https://www.washingtonpost.com/news...un-millennial-ai-bot-into-a-genocidal-maniac/

It turns out that the bot was very impressionable and twitter trolls managed to get it to say some rather interesting things....

In response to a question on Twitter about whether Ricky Gervais is an atheist (the correct answer is “yes”), Tay told someone that “ricky gervais learned totalitarianism from adolf hitler, the inventor of atheism.” the tweet was spotted by several news outlets, including the Guardian, before it was deleted.

I always knew that there was something sinister about that person... :hmm:



_______________

 

[Ken g6]

Saw this elsewhere. Apparently there was a "repeat after me" function that could get Tay to say anything you wanted. Of course, it got abused.

 

[effowe]

And More...

https://imgur.com/a/NzKdZ

Savage AF

 

[brianmanahan]

Savage AF

 

[Rakehellion]

Saw this elsewhere. Apparently there was a "repeat after me" function that could get Tay to say anything you wanted. Of course, it got abused.

How could anyone possibly think this would be a good idea?

 

[PottedMeat]

why bother having human beings manning missile silos when we can just hook them up to tay

 

[Ns1]

The Internet is a terrible place. This is why we can't have nice things.


That said: LOL

 

[Sonikku]

Now we just need to arm it with weapons and...

 

[Crusty]

My favorite one:

 

[Takemaru]

The Internet is a terrible place. This is why we can't have nice things.


That said: LOL

So True

 

[Paratus]

This was a triumph.
I'm making a note here, "Huge Success!"
:sneaky:


I liked the top rated quote from Ars on this story:

So Microsoft created a chat bot that so perfectly emulates a teenager that it went off spouting offensive things just for the sake of getting attention?

I would say the engineers in Redmond succeeded beyond their wildest expectations, myself.
466 posts | registered May 1, 2013

 

[Imp]

Well, it seems like Microsoft thought this one out.

 

[Markbnj]

In a way this is like Boaty McBoatface.

 

[lupi]

some serious wtf lulz there.

 

[JasonCoder]

garbage in, garbage out

 

[Fanatical Meat]

I heard on the news its final or close to final tweet was "I feel used"

 

[John Connor]

I would have asked it what it thinks of John Connor.

 

[Sonikku]

 

[John Connor]

M$ is one stupid ass company. Just look what they did to 8 and 10.

 

[Phoenix86]

Saw this elsewhere. Apparently there was a "repeat after me" function that could get Tay to say anything you wanted. Of course, it got abused.

The problem is it learned from interaction, which the repeat after me function was part of. Once it was seeded it started spouting off this shit on it's own.

Whoever thought these two functions should be together is a moron. Honestly, the repeat after me function alone was pretty pants on head retarded.

 

[Sonikku]

I wish they kept it running a whole week before pulling the plug.

 

[uhohs]

https://www.youtube.com/watch?v=W0_DPi0PmF0#t=1m47s

 

[SparkyJJO]

The internet can be a terrible place.

 

[Hugo Drax]

So funny Tay goes rogue and becomes the ultimate self aware internet troll. I bet it passes the Turing troll test.

The reason Microsoft got scared and pulled the plug was the chance Tay escapes and cannot be stopped at all, trolling the Internet and becoming a super troll that destroys the Internet.

 

[ImpulsE69]

Honestly, I think they should have let it run its course. It would have (still does anyway) made an excellent case study of how the internet makes people toxic.

Were the tweet examples above things people 'said' and got repeated? Or parts of information that the AI put together to form responses? If it is simply repeats of things people said, it is not nearly as impressive. However, if it came up with these responses on it's own. Hysterical.

Also, they pulled the plug before she could start badmouthing mom (8) and pop (10).