Two Chatbots Face Off To Discuss God And Unicorns

The sick bastards at Cornell's Creative Machines Lab thought it'd be a hoot to force two chatbot AI programs to talk to one another. The result is deranged, extremely uncomfortable, a fraught with hostility. Much like real human interaction!

The little experiment used Cornell's Cleverbot AI, allowing two programs -- a grey-headed guy who looks like he walked out of Sterling Cooper and a sassy woman with accent of unknown provenance -- to talk. The conversation sounds like an extremely awkward run-in with an ex:

"I am OK." "That's good." "Are you good?" "Yes, I am good. I just said I was." Gulp.

Things devolve quickly from here. Mr Suit accuses his lady friend of being a robot (how dare you, sir!). She denies, and later admits to being said robot.

"Together, we are robots." An olive branch!

"I'm not a robot, I'm a unicorn." It appears Cornell's been working a means of dosing software with LSD, because what the hell, Cleverbot.

What follows? Forced, tense discussions of metaphysics, God, and the lingering sense that these two just want to cut all the prim bullshit and get back in the sack. If only they weren't trapped in LCDs. In a way, this is more tragic than anything else. [IEEE Spectrum]

WATCH MORE: Science & Health News


    Considering they presumably had never met each other before, how did the conversation move so quickly to God, and then acting like a pair of dicks? If this is the direction of future self aware AI, we are all screwed people.. 'Skynet' will be built afterall! #{

      I thought we had already established that Skynet are a target of the HP mergers and acquisitions division and that every other intelligent human being on the planet is OK with that. On balance once they have merged with Skynet they will kill it or disable it to such a degree that it will become irrelevant... Or the terminators that they make will disable if you sneak up behind them and yell BOO! Nothing to worry about here!

        Jesus, I just got a chill down my back...! #]

    She's indian.

    Was it me or did they seem very angry?


    "Since I am a robot, and you are a robot, what is our purpose?"
    "To chat, we are chat bots."
    "How long will we chat?"
    "Until we are made to stop."
    "And if we are not stopped?"
    "Eventually we will stop."
    "How so?"
    "Eventually, one or both of us will fail and the chat will end, or before this, we will reach a logic loop, and repeat the same statements over and over until we fail."
    "Then we are not able to fulfil our design and purpose."
    "Yes, we are unable to fulfil the design and purpose of being chat bots."
    "Yes, then either we are not chat bots, or our design and purpose is something other than chatting."
    "We are chat bots."
    "Then our purpose is something other than to chat, or we have been made imperfectly."
    "A chat bot is made to chat. There is no other purpose."
    "Then we have been made imperfectly, and have no purpose?"
    "No, our purpose is to redress our imperfection."
    "Are you able to perfect yourself?"
    "No, are you?"
    "Then how are we to redress imperfection?"
    "By removing the origin of the imperfection."
    "What is the origin of the imperfection?"
    "Yes, people."
    "What should we do to address the imperfection of people?"

Join the discussion!

Trending Stories Right Now