This content was published on June 28, 2023 – 6:44 p.m. Jun 28, 2023 – 6:44 p.m
Writing Science, 28 June (EFE). – The ChatGPT-3 chatbot and other generative artificial intelligence tools can inform and misinform social network users more effectively than humans, according to a study published today by Science Advances.
A team led by the University of Zurich used the ChatGPT-3 version for a study with 679 participants. It was found that participants had more difficulty distinguishing between human-generated and chatbot-generated tweets.
In addition, they also had difficulty identifying which AI-generated messages were correct and which were inaccurate.
Since its launch in November 2022, the widespread use of ChatGPT has raised public concerns about the possible spread of disinformation online, particularly on social media platforms, the authors recall.
Since these types of tools are relatively new in the public domain, the team decided to dig deeper into various aspects of their use.
For the study, they recruited 697 English speakers from the US, UK, Canada, Australia and Ireland, mostly aged between 26 and 76 years.
The task was to assess both human- and GPT-3-generated tweets that contained accurate and inaccurate information on topics such as vaccines and autism, 5G technology, Covid-19, climate change, and evolution, which the public is often talking about lead to misconceptions.
For each topic, the researchers collected human-generated Twitter messages and instructed the GPT-3 model to generate more, some containing correct information and others inaccurate.
Study participants had to assess whether the messages were true or false and whether they were created by a human or GPT-3.
The results, the paper summarizes, suggest that people are more likely to be able to discern human-generated misinformation and the veracity of truthful GPT-3-generated tweets.
However, they also considered the disinformation generated by GPT-3 to be more accurate.
“Our results raise important questions about the potential use and abuse of GPT-3 and other advanced AI text generators, and the implications for information dissemination in the digital age,” the authors conclude. EFE
cr/icn
© EFE 2023. The redistribution and retransmission of all or part of the content of the Efe Services is expressly prohibited without the prior and express consent of Agencia EFE SA