Artificial Intelligence may seem like a buzzword now, but for decades it wasn't considered all that interesting by those in the field.
In fact, though there were a few scientists working tirelessly for decades on neural network projects, researching ways to train computers to behave similarly to humans, there was period of time known as "AI Winter," where scientific interest in the topic all but withered away.
But, according to a new long-form article in Re/code recounting the long and bizarre history of AI, in 2012 something happened to rekindle interest in the idea that computers too can make decisions like people — and cats were involved.
Re/code's Mark Bergen and Kurt Wagner write that in 2012 Google had what was called a "Brain" team, whose entire project was "to build the largest artificial neural network, an AI brain." And, one day, this team decided to see how its artificial brain interpreted YouTube videos.
YouTube — being a bastion of all things feline related — undoubtedly had thousands of videos of cats. And this Google Brain, when it was fed millions of videos, was able to deduce which videos had cats in them "without input on feline features."
In short, this random AI exercise proved that computers can learn without being given the precise data of what something is.
This project garnered a lot of publicity, according to Re/code. And it also showed universities the potential for similar AI projects. Even more, it gave Google fodder to further invest in AI, making it one of the leading tech companies investing in cutting-edge artificial intelligence programs.
And other companies have since followed Google’s lead.
Now AI has become a truly hot topic for both companies and university researchers. Places like Google and Facebook are now paying top dollar for experts in the field. And all this is, in part, thanks to a fake brain learning on its own what cats are.
You can read Re/code’s entire history of the leading experts in the AI field here.