1er With April approaching, consumers should sharpen their critical thinking when faced with certain advertising campaigns that could be designed by an artificial intelligence that loves April fools.
Experts are warning against the greater use of artificial intelligence (AI) in advertising because it raises new questions about the authenticity of a story. It becomes even more difficult to distinguish between facts, hyperfakes and even… April Fools.
Vivek Astvansh, a marketing professor at McGill University, mentions that many advertisements have been designed by artificial intelligence software like GPT-4.
While some April Fools’ ads are greeted with a smile, others create confusion or can even anger consumers. These promotional strategies represent a risk that can convince the population to turn their back on a brand rather than attracting them towards it.
Companies didn’t wait for AI to make jokes
For example, Google announced the 1er April 2019 that she had found a way to communicate with the tulips in their own language, “tulipish”. We could suspect the joke when we learned that this “translation tool” would only be available during this day.
Nowadays, greater accessibility to AI could allow more companies to make their own jokes of the 1er april.
“GPT-4 can instantly create content for many advertising campaigns. This can be videos or photographs. And in a very short period of time, while spending little money, the internal marketing team can look at what the software has created, make their choices and make some changes before running the ad,” says P.r Astvansh.
According to him, the key to avoiding being lured is by examining the methods used and the intentions of the advertiser, particularly on 1er april.
I hope that companies will inform consumers that advertising content was created or designed by artificial intelligence.
Vivek Astvansh, professor of marketing at McGill University
Digital watermarking is one of the ways that makes it possible to distinguish real images from fake ones while making it possible to identify the owners of this content.
“It is to ensure that the images or videos produced by these platforms are marked in such a way that we can determine their origin, that they were designed by artificial intelligence as soon as we see them on the internet,” says Sam Andrey, the executive director of Dais, a think tank at the University of Metropolitan Toronto.
The potential of AI to deceive an audience has already been well established. For example, fraudulent programs can perfectly imitate the voice of a loved one or even a politician. If we add the image of a well-known personality, the risk of misleading consumers is very great.
“Just a year ago, the images were more caricatured,” underlines Mr. Audrey, speaking of the progress made by technology.
He is not worried about the future if AI is used to produce innocuous images, citing the example of successful April Fools last year, such as Tim Horton’s square Timbits, the meatball vending machines meat from IKEA or Jeeps upholstered in flannel.
“But we should not use AI to deceive people,” judges Mr. Andrey.