Photo: Leon Neal/Getty Images
By: The opinion
The Bots have already become a part of the users of social networks.s, especially from Twitter; however, thanks to rapid advances in artificial intelligence, It’s getting harder and harder to tell them apart..
On May 4, 2022, Twitter assured that around 5% of its active users are bots. This platform reported 229 million users in the first quarter of 2022, which would place bots in more than 11.4 million botsonly in that social network.
Researchers from the Copenhagen Business School decided to conduct an experiment with 375 participants to check the difficulty of distinguishing between real and fake profiles on social networks
The researchers created their own mock Twitter feed in which the topic was the war in Ukraine. The feed included real and tweet-generated profiles supporting both sides. The fake profiles used synthetic profile images generated by computer with StyleGAN and messages generated by GPT-3, the same language model behind ChatGPT.
After the experiment, the experts found that participants were unable to differentiate between fake Twitter accounts artificially generated and real ones and, in fact, they perceived that fake accounts were less likely to be fake than authentic ones.
“Interestingly, the most controversial accounts in questions of accuracy and probability belonged to authentic humans. One of the real profiles was mislabeled as fake by 41.5% of the participants who viewed it. Meanwhile, one of the top performing fake profiles was only labeled a bot by 10%“, declared Sippo Rossi, PhD at the Center for Business Data Analysis in the Department of Digitization at Copenhagen Business School.
“Our results suggest that the technology to create fake profiles has advanced to such an extent that it is difficult to tell them apart of the real profiles”, added Rossi.
Bot proliferation
Another of the authors of the research considers that the growth of bots has been recorded in recent years because there has been a technological leap that facilitates the creation of this type of robotized profiles.
“Before, it was a lot of work to create realistic fake profiles. Five years ago, the average user did not have the technology to create fake profiles on this scale.the and with this ease. Today it is very accessible and within the reach of many, not just a few,” says co-author Raghava Rao Mukkamala, Director of the Center for Business Data Analysis at the Department of Digitalization at Copenhagen Business School.
From political manipulation to disinformation, through cyberbullying and cybercrime, the proliferation of social media profiles generated by deep learning has major implications for society and democracy as a whole.
“The authoritarian governments are flooding social media with seemingly sympathetic people to manipulate informationso it is essential to carefully consider the potential consequences of these technologies and work to mitigate these negative impacts,” adds Raghava Rao Mukkamala.
Keep reading:
• Twitter lied about the number of bot accounts and the security of the platform: everything that a former security chief of the social network said
• The personal information of 400 million Twitter accounts was leaked to the Internet: Donald Trump and AOC are among those affected
• Rise of artificial intelligence could lead to new religions that worship it