New Study Shows Gender Bias in Online Images Is Much Worse Than in Text
It’s even more pronounced than the disparities found in the offline world
3.2 billion. That’s the number of images being uploaded to the internet every single day.
In 2014, that number sat at 1.8 billion.
That’s a nearly twofold increase in just a decade.
And while it’s difficult to estimate exactly how many pictures in total there are online today, it’s safe to say this number is likely in hundreds — or perhaps even thousands — of billions and will only continue growing in the coming years.
After all, it’s not just humans who can create images anymore. It’s also the many AI-powered generative tools — like DALL-E, Craiyon, or Midjourney — that allow users to quickly and easily generate pictures using textual prompts. Plus, studies show that each year, we spend less and less time reading and more time consuming visual content.
But despite our online diets becoming increasingly filled with picture-packed news sites and social media platforms, much of the research into gender bias and stereotyping online has focused almost exclusively on text.
A recent major study aimed to fill this gap in the literature, though, by comparing…
