a thoughtful web.
Good ideas and conversation. No ads, no tracking.   Login or Take a Tour!
comment by user-inactivated
user-inactivated  ·  2946 days ago  ·  link  ·    ·  parent  ·  post: Microsoft brought its Twitter AI back online only to see it rapidly devolve into a drug abusing spambot

Probably to confuse old people who don't understand trolls. (To be fair, sometimes I fall under the traps of trolls too. It's a human thing.)

In all seriousness? I don't know. Maybe they're doing it as an outreach thing? Maybe they're trying to make an algorithm that will replace jobs like online customer support? If it was just an every day programmer doing this, I'd say they'd be doing it just to see if they could. Since it's Microsoft though, I'd guess they have an end game. The article I posted says they have a successful chatbot in Asia, so this isn't their first rodeo. Why they're having so much trouble with this one, I dunno.





oyster  ·  2946 days ago  ·  link  ·  

    Why they're having so much trouble with this one, I dunno.

God damn millenials ruin everything ! I think maybe they are trying to figure out a way to make a chatbot which is immune to this sort of thing and they figured the best way to test it was Twitter users. That's my best guess as to why anybody would think this was a good idea.

user-inactivated  ·  2946 days ago  ·  link  ·  

I think I read in a press release somewhere that they built a lot of fail safes into the bot, but for some reason they didn't forsee this particular scenario play out. I don't know a lot about these things, but I find that a bit hard to believe. I mean, 4chan loves pulling shit like this.

user-inactivated  ·  2946 days ago  ·  link  ·  

Since it's Microsoft I'd say someone wanted to play with chatbots on the clock, convinced Marketing a twitter bot would be a good idea to justify it.

user-inactivated  ·  2946 days ago  ·  link  ·  

"So your resume looks really impressive and you've had years of experience in the industry. Tell me, why did you leave Microsoft?"

"Well, apparently you're not looked upon too kindly when you create a Trump supporting, holocaust denying, toke machine spammer of a chatbot and let it loose on Twitter."

"Oh, dear. Did you really do that?"

"Well, no. But I didn't prevent it either . . ."