It’s easy to forgive Google’s early mistakes; most were made when the internet was in its infancy. But the company figured out quickly that there was more revenue in displaying official sites first. It knew, when you keyed in a bunch of ingredients, that you were looking for a recipe based on the contents of your fridge. These were easy fixes. Here’s a look at some of the bigger problems they’ve faced over the past 20 years, and failed to fix before they snowballed.
Accidental racism: In mid-2015, the fledgling facial-recognition software from Google Photos tagged two black teenagers in a picture as ‘Gorillas’. Google took the tags down in under 15 hours. That same year, it apologised again when a search for “unprofessional hairstyles for work” offered images of Black women, while a search for “professional” hairstyles yielded photos of salon-styled White women. It seems no one at the company headquarters realised that tech needed to be more neutral than its users.
Catalog Search: Between 2002 and 2009, Google had the bright idea to digitally record thousands of paper catalogues so buyers could browse online. It never took off. Why look at a paper brochure on the web when you can just look at a website? Google Answers met a similar fate when it turned out that Quora and social media answered questions better.
Google Video: A short-lived attempt to offer video services bombed. In 2005, two years after struggling against YouTube, it bought out its rival. The purchase angered plenty of content creators and users since YouTube had been slow to combat piracy or give creative people fair compensation.
Google Wave: In a Reddit AMA, Marissa Mayer, who was among Google’s first 20 employees and eventually headed Yahoo!, picked Wave as one of the company’s biggest mistakes. It was hyped as next-gen email in 2009, allowing multiple users to build a document through chat. Users were too confused to be impressed and the wave washed out.
Google Buzz: Among Google’s many attempts to compete with social media, Buzz integrated a whole bunch of products, including Twitter, in Gmail. Privacy infringement cases followed, and Buzz was discontinued in 2011. Then came Google+. No one uses that either.
Misreading AI: Google’s “photo-reading” algorithms seemed, at first, better at recognising objects and people than humans. But unlike humans, they ended up falling for optical illusions more easily and misread a figure of a gun as a turtle, or helicopter. Even as self-driving cars hit the road, tech is still being tweaked to overcome these hallucinations.
Fake news: Google’s tech, designed to allow anyone to upload material, didn’t account for the fact that enough users would publish enough lies to trick the algorithms into believing they were real. Years later, Google hasn’t done enough to produce better filters or take responsibility for its impact. Hoaxes and rumours now often spread faster than verified news reports, particularly in the wake of violent or traumatic events such as mass shootings, lynchings, floods or earthquakes. YouTube, owned by Google, has been criticised for offering loops of violent or inappropriate content even to child users.