Product reviews, deals and the latest tech news

Many people still don’t trust content created by AI, exclusively

Many individuals still aren’t sure when they’re confronted with AI-generated information, and they want to be informed whenever possible. When asked how frequently they had seen AI-produced material, a third of our respondents indicated they weren’t sure. More than half of our respondents stated they wanted content authored by AI (such as news, reviews, and features) should be explicitly labelled as such.

Vulnerability and openness

A further third reported being exposed to such material every day, and a fifth did so every month. Of those surveyed, 8.7% said they had not come across any material created by AI in the last several months.

When asked to define AI-produced material, the vast majority(84%) agreed that it must be either fully or mostly the product of AI. Just 9.5% of respondents agreed that AI assistance was not strictly necessary.

It’s understandable that people would like confirmation that the content they’ve been reading was indeed generated by AI, given the prevalence of such articles and the fact that popular AI tools like ChatGPT can produce text in a variety of styles that can be difficult to tell apart from human-created content.

Plus, artificial intelligence has been shown to make mistakes. Another reason why readers would want to know whether the text they’re looking at was created by a computer is because this might be the case. In the same way that people inherently evaluate the reliability of material based on its author, they will also evaluate the reliability of the information they are reading based on its author.

Recent experiments by popular tech news site CNET employing an AI engine to produce certain articles on the site have sparked irritation among readers since the site’s management hasn’t been entirely forthcoming about the matter. The articles it produced also had some elementary flaws.

The well-known chatbot ChatGPT performs no better in terms of accuracy. So far, it has made many blunders, such as providing incorrect geographical information (opens in new tab) and incorrect coding advise that got it permanently banned from Stack Overflow.

Only 7.5% of respondents said AI work was better to human authors and should be promoted, while 20% stated they were happy with it and didn’t care whether or not the creator was made public. Contrarily, 17% were strongly opposed, arguing that autogenerated material should be outlawed because it went too far.