AI and fitness tech are an ever-growing part of many people’s lifestyles, but the recent rise in “fitness AI model” like Aitana Lopez serve as a reminder that not everything as it seems. This opens up a debate on the potential for both beneficial and harmful applications for those who get their fitness advice online. M&F dives into the digital.
Would you know that Aitana Lopez was once a little baby pixel? Research is showing that it’s getting more difficult to separate AI from organic images. In one analysis, 260 participants attempted to label real people taken from Google searches, but were only able to get it right 61% of the time. Take into account that this study gave the subjects time to evaluate the images, whereas the fast-scrolling nature of viewing posts on a cell phone will make deciphering digital content even trickier.
@fit_aitana/Instagram
Fitness Model Aitana Lopez Isn’t All That She Seems
Once a figment of “The Clueless” modelling agencies imagination, the company now says that it is “Carefully curating thoughtful, long-lasting AI models that beautifully embody diversity,” according to the company’s Instagram blurb. Its hit creation, Atlana Lopez has amassed more than 300k followers on Instagram and reportedly earns her makers thousands of dollars per month. Lopez is part of a new wave of digital creations that are becoming more difficult to separate from real-life people. While previous generations of AI models were easy to spot, new techniques for making them appear real are taking the web by storm. Forbes reports that Lopez was signed as the face of a sports supplements company, while thousands of fans flood her with daily compliments. On Lopez’s IG page, she can be seen virtually working-out and looking super-fit in her preferred activewear, no doubt persuading her human followers to by the same brand.
“AI thrives on impeccable images, we do the opposite,” said the AI mode’s designer, Ruben Cruz in an interview with Marca.com. “For one, we don’t leave the skin so smooth. If we showed you a photo of how it comes out of the AI, you would know that that skin doesn’t exist.” To make the AI models more realistic, the far from ‘clueless’ developers upload flawed images into the software and use other techniques such as adding wrinkles and presenting Lopez with her hair out of place with an aim to lose those signs of perfection that often expose an AI creation. While Lopez’s origins as a line of computer code are no secret to many of her fans, not all digital companies are likely to be as transparent as Clueless, especially where there’s money to made.
The Use of an AI model in the Health and Fitness Space is Complex
The use of an AI model and the resulting positives or negatives within the fitness space is a complex issue that is unfolding at a rate many can’t keep up with. YouTube now hosts videos that are presented and voiced solely by AI characters, and this can make it difficult to trust the advice of someone (or something) who’s credentials cannot be verified. In the fitness the world, the notion of AI models churning out health advice based on an infantile or limited algorithm is another potential hazard. And yet the benefits of AI characters in terms of teaching humans, in a relatable way, are undoubtedly there to be harnessed. Another recent study found that while our engagement with fitness trackers and wearable tech is high, our ability to follow the guidance that they spit out often pales in comparison to those who are coached by a human being. The study’s authors theorized that in the future, superior generations of digital fitness wearables might include more convincing human AI. Considering the believability of Aitana Lopez as a real-life human being, that kind of progress is getting ever closer.
“One day, a well-known Latin American actor texted to ask her out,” said Cruz, speaking to Euro News. “This actor has about 5 million followers and some of our team watched his TV series when they were kids .He had no idea Aitana didn’t exist.” There’s no doubt that AI is an unstoppable force, but the key to getting the most from this technology rather than letting spurious deep fakes change the narrative will lie in how AI images and videos are regulated. There needs to be clear classification and verification if we are to ever trust out techie friends, no matter how beautiful they appear.