Artificial Intelligence is starting to play a big role in our daily life, and judging by recent developments in the field, it looks like the importance of AI will only increase in time. While some people worry that advancements in AI could replace jobs in the long run, it’s difficult to argue with the idea that artificial intelligence can bring countless benefits in numerous fields, ranging from production to automotive and healthcare. Needless to say, companies have different takes on machine learning, and artificial intelligence is used in more than one way. In Facebook’s case, AI will start playing a role in making social media interaction more meaningful to blind and visually impaired Facebook users who, starting today, can use a new feature called “automatic alternative text” in order to get a more detailed description of Facebook photos.
One of the main reasons why Facebook enjoys as much popularity as it does today is because of photo sharing. It’s been more than 10 years since Facebook users have been able to upload their photos on the social network, and since then the idea of photo sharing has been the basis of online social interaction. With this being said, it becomes quite obvious that a large portion of Facebook users is missing out on one of the main features provided by the social network. For countless Facebook users suffering from blindness or visual imparity, photo sharing and viewing is a feature that’s out of their reach. Fortunately, Facebook’s accessibility team has been working on ways to making photos relevant to this particular user base, and apparently it achieved this by using artificial intelligence to create a new feature called "automatic alternative text". In a nutshell, the feature uses machine learning to identify the subjects in a picture, and then uses VoiceOver on an iPhone to read photo descriptions out loud (in English) and to provide a more detailed explanation in regards to what the picture is all about.
Facebook’s new automatic alternative text technology can identify objects from various categories including nature, food, sports, transportation, and can even describe people through tags such as “smiling”, “beard” and “selfie”. The feature is available in the US, UK, Australia, Canada, and New Zealand on iOS starting today, and will also be released for Android and the web at an unspecified later date.