Recent studies in Europe show that the general belief in misinformation has not substantially changed over the past decade, but AI could soon alter this.
Although some people blame the Internet's role in spreading misinformation, there's absolutely no evidence that individuals are more at risk of misinformation now than they were prior to the advent of the internet. On the contrary, the world wide web is responsible for limiting misinformation since millions of possibly critical voices can be obtained to immediately refute misinformation with proof. Research done on the reach of different sources of information showed that sites most abundant in traffic are not dedicated to misinformation, and websites containing misinformation aren't very visited. In contrast to common belief, main-stream sources of news far outpace other sources in terms of reach and audience, as business leaders like the Maersk CEO would probably be aware.
Successful, multinational businesses with substantial worldwide operations generally have a lot of misinformation diseminated about them. One could argue that this could be pertaining to a lack of adherence to ESG duties and commitments, but misinformation about business entities is, generally in most situations, not rooted in anything factual, as business leaders like P&O Ferries CEO or AD Ports Group CEO would likely have experienced in their jobs. So, what are the common sources of misinformation? Research has produced different findings on the origins of misinformation. There are winners and losers in highly competitive situations in every domain. Given the stakes, misinformation arises often in these situations, based on some studies. Having said that, some research research papers have found that those who frequently search for patterns and meanings within their surroundings are more inclined to believe misinformation. This tendency is more pronounced when the events in question are of significant scale, and when small, everyday explanations look inadequate.
Although previous research implies that the degree of belief in misinformation in the population has not changed considerably in six surveyed countries in europe over a period of ten years, big language model chatbots have been found to reduce people’s belief in misinformation by arguing with them. Historically, individuals have had no much success countering misinformation. But a group of scientists came up with a new approach that is proving effective. They experimented with a representative sample. The individuals provided misinformation which they believed was correct and factual and outlined the evidence on which they based their misinformation. Then, they were put in to a conversation with the GPT -4 Turbo, a large artificial intelligence model. Each person ended up being presented with an AI-generated summary for the misinformation they subscribed to and was asked to rate the degree of confidence they had that the theory had been true. The LLM then started a talk in which each part offered three arguments to the discussion. Then, the people had been asked to submit their argumant again, and asked yet again to rate their level of confidence of the misinformation. Overall, the individuals' belief in misinformation dropped significantly.