In this day and age, AI is far too safe. No, I’m not talking about autonomous weapons – those are a real concern – I’m talking about social automation. Organizations and brands are releasing chatbots that are wholesome, kind, polite, accommodating, and completely fucking boring. The problem is that humans aren’t nice all the time. Sometimes we’re rude, aggressive, opinionated, and offensive. And if we are trying to create social interactions with automation that feel more human, more intuitive, and more authentic, it means we need to be willing to explore the full gamut of human socialization.

This was my motivation for my latest project, that I’m “beta-releasing” today (whatever the hell that means). Or, at bare minimum, this was my excuse for creating a hilarious, sociopathic chatbot named Shananigans. I won’t bore you with the nitty gritty of how it got done (other than to say that Matt Read is my bloody hero) but I decided to publish Shananigans below for two reasons. First, I wanted to run a social experiment with people to gauge their thoughts and reactions when Alexa, Siri, Google, or Cortana stop playing so nice and have a bit of teeth. Second, I think a chatbot swearing at you is hilarious and I wanted to share my asinine humor with the public.

So play below and reach out to me if you have any thoughts, questions, or comments. Also, while the text version works well below, if you’d like to interact over audio (Android Phone, Google Assistant, or Google Home), shoot me a line and I can add you as a reviewer to the project and teach you how to raise an instance of the chatbot app to your device.

https://bot.dialogflow.com/shananigans

Shane Saunderson plays with robots, minds, and organizations.