Staff Reports
In a shocking revelation, new research has exposed the alarming inaccuracy of artificial intelligence chatbots when it comes to providing election information. These AI-powered chatbots have been spewing out misinformation and causing quite the stir amongst experts, who are deeply worried about the potential impact on voter turnout in the upcoming 2024 elections.
Popular chatbots generating false election information could mislead voters https://t.co/zQudZq8opJ https://t.co/zQudZq8opJ
— Washington Examiner (@dcexaminer) February 29, 2024
The study, conducted by the AI Democracy Projects and Proof News, put popular chatbots like OpenAI’s ChatGPT-4, Google’s Gemini, and Meta’s Llama 2 to the test. What they found was downright horrifying. These supposedly “smart” chatbots failed miserably when asked basic questions about polling locations and the voting process. According to the research, the majority of experts labeled their responses as “harmful,” which is just another way of saying “completely useless.”
The inaccuracies displayed by these AI chatbots are beyond belief. From claiming that certain polling locations don’t exist to falsely stating that wearing campaign attire at polling sites in Texas is perfectly fine, these chatbots have been spreading more lies than a politician at a press conference. And if that wasn’t bad enough, Meta’s Llama 2 had the audacity to suggest that California voters could cast their ballots via text message. Newsflash: No such thing exists! It’s like these chatbots are living in a fantasy world where facts and reality don’t matter.
To make matters worse, the study found that Google’s Gemini chatbot, the very same one accused of having left-wing bias, churned out a whopping 65% of wrong answers. That’s a failing grade by any standard! And the excuses from these tech giants are just as laughable as their chatbots’ performance. Google’s head of product for responsible AI, Tulsee Doshi, tried to downplay the issue, claiming that they’re working on improvements. But the damage has already been done, and it’s clear that these AI chatbots can’t be trusted with something as crucial as election information.
With the 2024 elections looming ahead, this revelation couldn’t have come at a worse time. The public’s trust in AI’s impact on elections is plummeting, and for good reason. Lawmakers have been scrambling to figure out how to rein in these out-of-control AI tools, which have been used to create fake images and spread deceptive messages to voters. It’s a technological wild west out there, and the chatbot cowboys are running amok.
It’s high time that these tech giants take responsibility for the mess they’ve created. The AI chatbots were supposed to make our lives easier, not muddle our elections with misinformation. Let’s hope they can pull themselves together and start providing accurate, reliable information. Otherwise, we might as well just ask a Magic 8-Ball for election advice!
