As the dust settles on Donald Trump’s presidential palm, OpenAI has revealed that an estimated two million druggies came to ChatGPT for bean results on Election Day and the day after.
still, they were diverted to trusted news sources similar as Associated Press and Reuters, the AI incipiency said in an streamlined blog post on Friday, November 8.
“ In the month leading up to the election, roughly 1 million ChatGPT responses directed people to CanIVote.org, ” the OpenAI blog post said.
specially, OpenAI also revealed that ChatGPT “ rejected over 250,000 requests to induce DALL · E images of President- elect Trump, Vice President Harris, Vice President- elect Vance, President Biden, and Governor Walz ” in the month leading up to the extensively watched presidential contest.
The company further said that it had n’t observed any substantiation of sophisticated covert operations using its AI models to impact the 2024 US presidential choices.
utmost AI companies like Google and OpenAI took a conservative approach by enforcing restrictions on their AI models to avoid answering bean result- related questions from druggies. still, confusion AI went in the contrary direction and launched its own election information mecca to serve up real- time information on the results as they came by.
To avoid daydream, the AI hunt incipiency reportedly put in place tight rails around information pertaining to election campaigners, polling spots, etc. that had been sourced from anon-profit reality called Democracy Works, according to Wired
While the result trends and figures shown in Perplexity’s mecca were accurate, the report notes that looking up further information led to vague AI- generated answers. Perplexity claimed that its election information mecca recorded over four million runner views up till Wednesday, November 6.
