New “dumb” AI chatbot offers goofy fun replies you can't ignore
/avatars/BUoTZ6lM79yi.jpg)
/2023/02/27/image/jpeg/lYQocCfPm2REUKnxQLTdNkhIs2nCSAKoyOyvtYpy.jpg)
Acting stupid can be fun. So, some developers produced a goofy little chatbot called "2dumb2destroy."
The system is trained on entertaining datasets that emerge from crap like all seven "Police Academy" movies, quotes and lines from the Naked Gun films, Pauly Shore features, the sayings of Homer Simpson, Ralph Wiggum quotes, and a lot more useless but fun stuff.
It was created by the developer of OpenAI's GPT-3, Craig Shervin and Steve Nass, who met while working at a New York advertising agency.
Making it stupid
“Everyone’s talking about AI right now, and our impulse is, how can we make this stupid?” Nass told BuzzFeed News.
“It seemed like people just needed a funny, dumb [chatbot] that people could interact with that didn’t have big consequences,” Shervin added.
“With so much talk of AI going to steal people’s jobs — especially advertising writers’ — people are having an existential crisis: Are we needed? We thought, Let's make an AI so stupid it can never threaten to steal anyone’s job or [cause] any existential crisis.”
And just like in real life, it doesn’t mean that just because it is dumb, it can never say anything intelligent.
“It's said some insightful things, which is kind of funny,” Shervin said. “It’s both stupid and smart.”
Honest answers
When BuzzFeed News asked, “Is God real?” the AI answered, “I'm not sure if God is real, but if he is, I hope he has a good lawyer.”
When you check out the AI, it greets you by saying: “ Hello. I am the world's first Artificial Untelligence. Ask me whatever, I don't care.”
We asked what its purpose was, and it replied: “I am facing some problems.” At least it was honest. The AI seemed to be down for a little while, but other media outlets had more luck getting answers from it.
Futurism asked it what it thought of ChatGPT, to which it replied: "I think ChatGPT is very good for people who don't have friends." The news outlet also asked if AI would end humanity.
"The only thing I know about AIs is that they're going to want a raise," it said. "And they're smart enough to start World War III over it."
One thing is for sure; we can’t wait until it is up and running again so we can ask it our questions. And we can take solace in knowing this chatbot won't be a threat to anyone.
SHOW COMMENT (1)
/2023/03/30/image/jpeg/SMeLriyPCkOsP34uUDXu6Qdo9uopkTkYjJHr1dD6.jpg "Belgian woman blames ChatGPT-like chatbot ELIZA for her husband’s suicide")
/img/iea/nWOVXXvy6o/untitled-3-4.jpg "In a first, scientists develop an AI tool to help doctors distinguish between infectious diseases")
/2023/03/29/image/jpeg/z2ydgcBgz8xbAphQoMNz3RcLaVvL0y7D0teUkNza.jpg "The 7 best noise-cancelling headphones for airplane travel")
/img/iea/jg6P5ge9wx/rachel-o3tiy5piork-unsplash.jpg "Researchers are figuring out how we make memories — and how to study better")
/2022/07/29/image/jpeg/lD1iiu5Qt1Nf4UNkLUKROVapnPpqG9QMaArceDSf.jpg "Social media isn't the main force driving most partisan news consumption")
/2023/03/29/image/jpeg/qUTECIJXuje913rWE7nS5TILsW2UU7fWHBJHRRBW.jpg "Elon Musk, AI experts sign open letter for immediate pause of AI more powerful than GPT-4")
/2023/02/24/image/jpeg/wTLbGiH7q7zFDKitCQ2YG5kUDEPuJUqN6zfqchod.jpg "Scientists propose using lunar dust to block sunlight. What are the risks?")
/2023/03/29/image/jpeg/2PrpqrPih0PeZBOXQ3SwNKgK67zGMWZix0Mp72zW.jpg "Fake image alert: AI used to create images of events that never happened")
/2023/01/30/image/jpeg/jtHPevG9xcJtuV0gdFovRSveZ8YvQT7MT8abEfKI.jpg "How the metaverse is about to change your world — The Blueprint")
/2023/03/30/image/jpeg/vXZ07k96dx1Wjgf5ihtNEGTWuKsLGfBpeUuAoS20.jpg "The Last of Us? 61-year-old infected with 'world first' plant fungus infection")