It was inevitable. The ChatGPT software was used in a political campaign by a bot developer working for Rep. Dean Phillips.
Silicon Valley entrepreneurs Matt Krisiloff, Jed Somers, and others created Dean Bot to provide interactive information for Phillips. They created a SuperPac called We Deserve Better ahead of next Tuesday’s New Hampshire primary.
The SuperPAC contracted AI startup Delphi to build the robot. OpenAI closed Delphi’s account on Friday night after the Washington Post published a story about the SuperPAC that revealed the existence of the bot.
OpenAI spokesperson Lindsey Held stated in a press release that “anyone who builds using our tools must adhere to our usage policies.” “We removed a developer’s account recently that violated our API usage policy which prohibits political campaigning or impersonating a person without their consent.”
Washington Post:
Krisiloff told The Post that he asked Delphi to remove ChatGPT and instead use open-source technologies which also provide conversational capabilities. These were built into the bot design.
Krisiloff stated that the bot was available for public use without ChatGPT up until late Friday when Delphi removed the bot in response to its suspension.
It is commendable that Delphi has taken down the bot. It’s important to distinguish between what is real and what is artificial intelligence in the early stages when AI is mixed with politics.
The voters have no idea of what’s about to happen. Imagine a Donald Trump or Joe Biden that is fabricated to appear to make false or misleading statements about minorities, or other key constituencies. Today, the adage that “A lie can travel halfway around the world while the truth is putting on its shoes” is more relevant than ever.
The bot came with a disclaimer that explained that it was not Dean Phillips but an AI tool, and that voters had to consent to its usage. Researchers told The Post, however, that these technologies can lull users into accepting dangerous tools even when they are disclaimed.
We Deserve Better is one of the proponents who argue that bots can be used to educate voters in a fun way.
It is misleading to say “Used appropriately”. Who decides what is “appropriate?” The notion that bad actors don’t give a damn about the “appropriate” use of technology, and instead use it for their benefit or to promote a cause is absurd.
The AI genie has escaped the bottle. The AI genie is out of the bottle. It’s better to accept it than to try to stop the technology from bringing about incredible benefits.
