Facebook AI – Opensourcely Hoping To Be On Our Devices?

Facebook AI, keeping your data safe while reading your private messages.

Facebook and encryption, kind of an oxymoron right?
What is encryption? It keeps your data private. Well, that’s not exactly true.
If surveillance software is installed on the device prior to use, then encryption may as well be a megaphone. Being on the device means the software can monitor every keystroke. As you are typing that message it is readable on your screen, ergo, the surveillance software can see it.

This is, in essence, what Facebook are peddling with ‘AI On Edge’. It is called AI, and, it definitely looks at information with what could be considered intelligence. It then processes this information and reports it to whoever needs to see it.

So, you see the issue here. If Facebook manage to make this a reality, then encryption may as well not exist any longer.

Facebook are trying to create something that will analyse what you are typing before you hit send. This cannot happen without being installed on your device first. The issue will really come when manufacturers include this in their devices. This will mean nothing is private any longer. 

‘I have nothing to hide though.’

Yes, that may well be true. But what you consider nothing to hide, may seem otherwise from a strangers point of view. Look at it this way. If you feel strongly that you have nothing to hide. Would you be happy to shout out your messages to others in a busy street? Regardless of how personal the content is? We doubt it. Those strangers you would not want to air personal, possibly intimate information to, could just as easily be the people monitoring your chat.

Is it not just AI though that will perform this monitoring?

Yes, generally it will be. The problem will be when ‘checks’ have to be made. When a human being will have to view chats from supposedly anonymous users. The thing is, without encryption being exactly what it says on the tin, there is no ‘anonymous’.

Facebook AI – AI On Edge (on device)

Currently, Facebook use AI to check content as it is posted online. This stops people violating policy with text, images and video that are specifically posted to cause distress or revulsion. This is fine as the amount of content is really beyond a handful of humans monitoring. This current AI is not on your device.

If you watch the video below, you be the judge. It sounds very much like Facebook will be working towards making small chips with the AI embedded on them.

AI on device
(May 01 2019)

The video is a full in depth discussion on the challenges Facebook have in needing to know more information.

Skip forward to 23:00 minutes onward for Facebook’s specific ‘AI on Edge’ plans. We suspect the decision of calling it ‘AI on Device’ opposed to Edge may have sent the wrong message.
As in, a transparent one.

For more gadget and tech related information, check out gadgethelpline.