Presumably one of the major benefits of artificial intelligence is its ability to perform tasks faster and with less complaining than a human can. So it only makes sense that one of the first exercises to go extinct is the arduous task of tagging and sorting image files.
Box —a $2.48 billion content management company — has partnered with Google Cloud Vision to apply Google Images search technology to Box's storage technology. The feature launched in beta mode Thursday for free.
The feature, called Box Image Recognition, is opt-in. Enabling it means that Google has access to your Box account, but a Box representative said that data won't be cached and Google deletes images once they're analyzed.
It works by tagging images as they're uploaded, making swaths of visual data searchable with key terms. If you're looking for a photo of pants, for example, just type "pants" into the Box search bar, and voila.
Google Cloud Vision is so well trained from years of machine learning and enormous data sets that it can even tag with abstract phrases that capture the essence of a photo.
Rand Wacker, VP of Product Marketing at Box, said the technology can also read text.
A company could use it to automatically run a background check, Wacker suggested, with just a photograph of a driver's license. The AI can read the ID, and then automatically kick off the process defined for drivers licenses, if a company so chooses.
Longterm, Wacker said that Box intends to work with multiple other search partners for more robust capabilities. Different image recognition technologies are fine-tuned for different data, and might work better for some industries than others.
Box Image Recognition isn't without competitors. Amazon's Rekognition, for example, was trained by analyzing the billions of images uploaded daily into Prime Photos.