How Does Facebook Use AI?

Looking to Make Amends with the Help of an Improved AI Offering?

Whoever you are and wherever you live, it is reasonable to assume that you already know about Facebook’s misuse of user data.

What you may not know is that over recent months, more and more information has come to light related to Facebook’s development of its own Artificial Intelligence systems.

Judging a CNBC report on the subject, the driving feature of Facebook’s AI at this time is that it has been built on algorithms that have the ability to analyze 52,000 user traits. With this in mind, it would seem that they are now building an AI system to predict customer behavior in some detailed fashion, which one could argue indicates a fine line between a helpful development and another data scandal.

In attempting to truly understand why Facebook would develop such a technology in the face of such a possibility, it is essential to understand what their current business model is. In a general sense, any chance that they have of making a profit is based on selling user data to advertising companies, so that they can provide highly targeted advertisements to these companies on an almost user by user basis. It is exactly because of this that Facebook was called to testify in Congress and lost $100 billion in stock value.

They were selling too much user information, while hiding the provisions that supposedly allowed them to do so.

At this point, it is easy to understand why Facebook is reluctant to give up on its current business model, given that by all appearances, it has no other viable option in place.

In short, how else would it generate a profit?

Upon reading further in our primary source, it also becomes clear that Facebook might be attempting to brand its efforts as something new or unique, but they are nothing of the sort. Furthermore, by all accounts, it appears that Facebook aims to push forward even harder with its current business model, regardless of the scandals that it has already experienced. In the case of its AI, the only new development appears to be the number of AI algorithms that it will be using to continue its efforts. More specifically, they have apparently added three, which are called DeepText, DeepFace, and FBLearner Flow.

Each one has a different purpose and a different area to target.

All of them look like they are threats to the average user, at least on the surface.

For example, in the case of DeepText, it really looks as if it is a secretive mechanism to bring about a time in which consumers feel, but do not know if they are choosing what they search for. In truth, “feel” is the key word here, which transitions us well to our upcoming post on the subject.

For now, I leave you with this question: do we really have control over anything that happens under Facebook’s sphere of influence?