Pictures of drunk people on
Facebook may lessen in future, if plans revealed by the social network's AI research lab come to fruition. Lab head Yann LeCun wants to create a digital assistant for Facebook which is capable of detecting whether photographs being uploaded to the service are being done while the user is under the influence of alcohol, with the aim of preventing anything the user would find embarrassing when they are sober from being shown to visitors.
According
to Wired, the plan is already being worked on by the team in offices located in New York and California, with the effort centered a "deep learning" AI which researchers hope will be able to tell if the user is sober or drunk at the time the picture was taken. Ultimately, the system could operate by giving the user a warning before an image is made public, asking "Are you sure you want your boss and your mother to see this?" "Imagine that you had an intelligent digital assistant which would mediate your interactions with your friends, and also with content on Facebook," suggests LeCun.
Some functions the project requires are already being used on Facebook. The social network is easily able to identify the user and friends' faces in posted photographs for easier user tagging, with deep learning algorithms already used to select content to show in the user's news feed. The same technology will also apparently suggest relevant hashtags for status messages, something which will apparently arrive on the service soon.
There is no word on when such a "drunk-posting protection" system will be ready for public use, if at all.