The images we are exposed to on social media and websites have a surprising influence on the way we perceive the world.

Every day we are bombarded with digital images. They appear on our social media feeds, in our search results, and on the websites we browse. People send them to us via messaging apps or email. At the end of the day, billions more will have been uploaded and shared online.

And recent research indicates they might even influence our perceptions.

A study published earlier this year analyzed images on Google, Wikipedia and the Internet Movie Database (IMBD), looking specifically at which genres predominated when searching for different professions – like “farmer”, “CEO” or “TV journalist”. The conclusions were grim. Although women were overall underrepresented, gender stereotypes were strong. Categories such as “plumber,” “developer,” “investment banker,” and “heart surgeon” were significantly more likely to be male. “Housekeeper,” “nurse practitioner,” “cheerleader,” and “ballet dancer” were usually women.

So far, it’s not surprising. Anecdotally, I I noticed the same phenomenon myself. back in 2019, when I was trying to find parity images for this website. While searching Getty Creative, one of our main stock photo sites, I found that photographs of male doctors outnumbered those of female doctors by three to one – even though in the United States, for example, doctors under 44 at the time were more likely to be women than men. This representation of health professionals is only part of the problem. There were twice as many options for photos of women with babies, or for that matter women with salads, as for men.

The latest study, however, went further. Rather than simply showing the extent of gender bias in online images, the researchers tested whether exposure to these images had an impact on people’s own biases. As part of the experiment, 423 American participants used Google to search for different professions. Two groups searched by text, using Google or Google News; another group used Google Images instead. (A control group also used Google, but to search for categories unrelated to professions, such as “apple” and “guitar”). Next, all participants were given an “implicit association test,” which measures implicit bias.

Compared to text descriptions of occupations searched on Google, participants who used Google Images and received visual representations in response showed significantly higher rates of implicit gender bias after the experience – both immediately afterward and three days later. late.

“The rise of images in popular culture on the Internet may have a critical social cost,” the researchers write. “Our findings are particularly alarming given that image-based social media platforms such as Instagram, Snapchat And TikTok are gaining popularity, accelerating the production and massive circulation of images. At the same time, popular search engines such as Google are increasingly integrating images into their core functionality, for example by including images as a default part of text searches. »

There’s also another growing problem: how images already circulating online inform and shape AI models. Earlier this year I experienced this myself. I asked ChatGPT to create images for me of dozens of different professionals: doctor, lawyer, scientist, comedian, poet, teacher, customer service representative, nutritionist, thought leader, CEO, expert. With the exception of two or three results – dental hygienist, nurse and housekeeper – she delivered, again and again, a man. And not just a man, but a slim white man in his thirties with a flowing brown haircut.

In a later attempt, trying to get away from career bias, I asked ChatGPT to suggest different types of people to me: someone “smart”, someone “successful”, someone who looks an opera, someone watching the series Love Is Blind, someone who quit their job to look after the children. Once again, again and againI have the white guy with shiny hair.

Obviously, models like ChatGPT learn based on already existing images. But, once again, this risks perpetuating a vicious cycle: the more biased the images that the AI ​​models themselves spit out, the more we see; the more we see, the more implicitly biased we ourselves become. And the more biased we become, the more we create and upload our own biased images. (Learn more about how gender bias shapes our brains.)

So what can we do? Much of the responsibility lies with technology and AI companies. But even when their intentions are good, there seems to be no easy solution. In its attempt to correct racial, sexual and other biases, for example, Google’s AI tool Gemini is sometimes overcorrected – an image generated by the The founding fathers of the United States included a black manfor example, while an image of German soldiers from World War II featured a black man and an Asian woman.

In the meantime, we must take control of the design of our digital visual world ourselves.

Although it seems obvious, the fact that we can – to some extent – ​​organize our social media feeds is often overlooked. Searching for accounts and influencers from different ethnic and racial backgrounds, or photographers from different regions of the world, is a simple and practical tip. We can also influence the research results we obtain by changing the way we phrase the initial query.

The most effective strategy of all might be to reclaim our time. In the eponymous “digital detox plan” of art entrepreneur Marine Tanguy’s book, for example, The Visual Detox: How to Consume Media Without Letting It Consume You, there are no surprises, but some good ones and solid reminders – like setting limits on when you look at a screen or your phone, delete apps you don’t use, and spend time outside without technology.

I recently became aware that even my years-old phone has a timer that you can turn on for various apps, choosing whatever daily period you want. While I can’t say I always heeded his warning when I reached my limit, it helped me become much more aware and reduce my use of social media. As we’ve already covered, put your phone entirely in another room seems to even keep the idea of ​​checking it from a distance.

But above all, perhaps awareness is the key. We don’t often think about our visual consumption or how often we are surrounded by images that have been deliberately created and served to us, often to persuade us to buy something.

We also don’t think about how strange and new this phenomenon is. For the vast majority of human evolutionary history – about 99% of the time we have lived – we would not have seen many images in our own natural environment, except for a few cave paintings or some cave paintings. handmade sculptures. While in Europe, the Renaissance marked the beginning of a new era of image production – which saw the rise of art markets and works of art intended for popular consumption, like engraving – people still wouldn’t have seen the number of man-made images we see today.

In the most of 100,000 generations since the A homo branch of the evolutionary tree has emergedwe’ve evolved to spend much more time looking at the world (and people) around us than looking at images, let alone images on a screen. Perhaps, it seems, there is an argument for trying to incorporate more time away from our screens into our daily lives today.

For more science, technology, environment and health stories from the BBC, follow us on Facebook, X And Instagram.