Microsoft certainly hasn't been scared off doing further experiments with artificial intelligence after the whole racist bot debacle.
The tech firm is back with an AI powered API that can identify (or at least try) what's in a picture.
1. Boat or wheel?
2. The shard clock
3. Unhappy Sir Martin Sorrell
4. Tech blind?
5. Apple fan
6. Dog lover
7. Cameron BFF
Created by Microsoft's Cognitive Services which was also behind the how old robot and the twin or not robot , CaptionBot uses computer vision and natural language to come up with descriptions of the image.
Test it out yourself, here (warning, the images you put in will be "held on to" by the bot to learn from).