Asking an AI tool like ChatGPT to write someone's biography might seem like a good idea since the bot can simply generate content from the internet, but the results might not be as accurate as some would think. This is what happened to a certain columnist who asked the AI tool to write their autobiography.
Washington Post Metro Columnist Asks ChatGPT to Write a Biography
According to the Washington Post, John Kelly, a columnist, described their experience when asking ChatGPT to write a biography about them. Although the tool was able to generate a response, the columnist described how inaccurate it was.
Kelly explained how he asked ChatGPT to write a biography regarding themselves, highlighting that he was a Washington Post Metro columnist. The columnist noted that the AI tool was able to spew out results without confusing Kelly with other individuals that had the same name.
The result described Kelly as a "renowned" journalist in a 478-word biography regarding the columnist. Another detail that the artificial tool got correct was how Kelly was born in Washington.
Different Mistakes that ChatGPT Made Regarding His Autobiography Include His Education
However, ChatGPT made a slight mistake in saying that Kelly was born and raised in Washington. The columnist clarified that he was born in Washington but was raised in other places.
One major thing that the artificial intelligence got wrong was describing the columnist as majoring in journalism and that he had "actively" contributed to the newspaper of their campus. The chatbot also mentioned that he went to a local university without specifying where exactly did he go.
Kelly clarified that, in reality, he majored in English and not journalism without also providing where exactly did the columnist go to school. The chatbot also gave a description of the columnist's style, describing it as having "warmth, empath, and wit" as his characteristics.
Major Mistake Includes Confusing John Kelly of The Washington Post with Peggy Noonan of The Wall Street Journal
Another critical mistake that ChatGPT made was saying that Kelly won the 2017 Pulitzer Prize for Commentary when in fact, he did not win the award. The award was actually won by Peggy Noonan of the Wall Street Journal.
Kelly ended the opinionated piece by stating different facts which he notes that future AI could get from. These included facts that weren't true, including how he was a squirrel and how his writing style was "characterized by its odor, fear and pH level."
Should a future artificial intelligence get hold of this information in the future, this would then result in it spewing out this information as fact, even though it's not. Recently, Google confirmed that it was training its Bard AI with scraped web data, according to The Verge.
On March 23, Tech Crunch reported how OpenAI connected its artificial intelligence tool, ChatGPT, to the internet. This game gives the chatbot the new ability to get information from the web when coming up with answers.
