It only took 7 years!

During 2010 I went on a trip to South Africa and while visiting Robben Island prison I took a photo that I used in a blog post about metadata.


At the time I mentioned that tagging of photos to make them easier to find later was a mostly manual task with only a timestamp and location being automatically captured. I also speculated that tagging would be automated some day…

Since your camera can’t determine (at least not yet) whether a photo contains a boat, you need to manually point out which photos do contain a boat…

And here we are seven years later and we have an API that can look at that same image and tell me that it’s “a large ship in a body of water”:


What’s even better is that you’re seeing this show up more and more as a just a feature in the apps we use every day like Office and Photos 😁