Thursday saw Social Networking giant Facebook making a significant announcement. The California based company shared its plans of developing ‘DeepText’, a new artificial intelligence (AI) system.
The initial information made available by the company revealed that the AI will have the ability to analyse what people are writing about in the Facebook posts with an accuracy that is almost near-human. In addition to this, it will be able to achieve this in across 20 languages.
According to a statement shared by Facebook with a technology website, if the AI is able to automatically pick out hate speech from Facebook status updates that merely contain controversial terms, it would help prevent a Facebook user from ever suffering the harassment and going the process of reporting it manually.
Facebook strongly believes the technology needs to be developed up to a level that computers are able to understand things like word-sense disambiguation and slangs etc. as this will help them get closer to how humans comprehend text.
Facebook has already started testing the DeepText on its Facebook Messenger and has got very promising results. According the networking giant, the AI systems have successfully reported more offensive photos than humans on its networking site.
According to Facebook, with DeepText its aim is to make optimum use of its ability to make better recommendations to Facebook users, both about related posts and related services one can avail on the network and improve its overall user experience.
For example, if the AI recognizes in a sentence that the user is in need of a taxi ride, it will suggest them to make use of Messenger’s transportation integrations with services such as Lyft and Uber.
Just when you think our Facebook experience can’t get any better, the networking giant comes out with something new and innovative.
[Top Image Credits: Michael Short — Bloomberg via Getty Images]