You’re being manipulated by Artificial Intelligence, and it’s not even sentient yet. Here’s why you should be concerned.

This article has been a long time coming.

Let me just clear the air a bit and cut to the chase to save you all a fairly long read. Artificial Intelligence doesn’t have to be thinking to control you. It doesn’t have to be ultra-intelligence, it doesn’t have to think like JARVIS from the Iron Man movies. It doesn’t have to be a near alien or god-like being.

All it needs is a task. A task and optimisation parameters.

You can end the article here, or you can inform yourself. If you do end it here, I hope you take the message above as a warning - if you keep reading, then by the time you wrap this article up, you’ll be able to discuss Artificial Intelligence with some degree of proficiency.

Oh, one more thing - yes, thinking AI which will pass the Turing test is definitely coming, but there’s no need to worry about that for the moment. Let’s focus on the now - because the present is what defines the future.

You’re being used. That’s all there is to it.

I’ve been in the field of Artificial Intelligence now for quite some time. Big Data, Data Science, and the rest of it have sort of been on the backburner as it were. It started off as an interest brought on by scifi concepts, videogames, and the rest.

In Google’s database, in Facebook’s as well, they have a file with your name on it or a profile set with a number attached to it and details of what you like, what you don’t like, the places you visit, the things you’re interested in, and the rest of the details you provided such as marital status, age, etc.

This isn’t all that spooky. We have heard this before.

You may be a touch freaked out thinking that deep in the servers of Google there’s a file with a number on it that matches you. And I understand why that may be worrying.

But why is Artificial Intelligence scary? Well for one, it can narrow you down to a tiny detail.

Consider for a moment, all the people who use Facebook and Instagram. There’s bound to be some overlap over multiple groups of people. For instance, if you like Japanese Food then there’s a chance you may like Anime. If you like Anime, there’s a chance you may like Manga.

So on and so forth. These are of course assumptions.

What having access to the data at Facebook allows you to do, is to make predictive assumptions based on that data. You can expand your particular target to attract more attention. You can use Facebook’s free affinity tool to figure out your demographic better and what they like.

It’s for this reason you see random products that you talked about not all that long ago. It’s Facebook tracking your location, where you’ve been, and a couple of other factors.

It’s also for this reason you see more content about a particular type of thing pop up on your news feed. Notice that if you like pictures of dogs, suddenly there are loads of dog posts? That Facebook telling you to engage more. It’s making you engage more.

Every now and then, Facebook will see what you’re engaging with and optimise. The logic behind the Artificial Intelligence would be something along the lines of:

“Well, okay, Ed over here is liking 6 dog pictures per day… Why don’t we put three pics of cats in front of him?”

If you like all of them, it assumes you like animals and dogs equally. Now there’s an influx of both.

If you like just the dogs, it’ll assume you dislike cats but it may repeat the test later.

If you like just one cat picture, it’ll assume you sort of like cats given a certain context.

Ah, but, if you tag someone in a cat picture - Facebook’s algorithm will begin to assume that the person you tagged likes cats, then will begin testing cat pictures on them.

This is how Facebook builds a constant profile of you. What are you engaging with? And more importantly, why? Then it’ll keep creating a better and clearer image.

Now that Facebook knows you prefer dogs over cats, it’ll put adverts in front of you which are related to dogs. You don’t seem to care about dog collars or dog food? Then you’re just a dog enthusiast. But if you click the ad, suddenly it knows you own a dog.

Facebook can then add this data further and make assumptions based on existing data from other users. Although this is probably not a true fact, let’s pretend that 76% of dog owners drink Coca-Cola.

The Facebook AI will now assume you might enjoy Coca-Cola given that 76% of people who own a dog do enjoy Coca-Cola. So on and so forth. Such are the intricacies of the Artificial Intelligence model on Facebook.

This can be used for serious nefarious purposes as you can probably tell. If someone were to look into the affinity rate of dog owners and a particular political party, they could target their adverts towards dog owners. Facebook’s targeting and marketing tools make it easier than ever to track down targets. All providing the affinity rate was high enough of course.

My greatest achievement was targeting 5 people who went to my University using very specific data and methods. All in the name of journalism of course.

This leads to manipulation of course.

Imagine having the ability to show a highly focussed group of 1000 people a message that has an 80% chance of eliciting an emotional response. A message which may be completely false and untrue. You could easily mislead thousands - and the only safeguard is Facebook hopefully analysing the content thoroughly before allowing it to be pushed through. Even then, it would only be marked with a “Potential Fake News” just below the video.

I realise that this leads to a lot of questions involving Artificial Intelligence and where we stand having all this data at our fingertips, with money as the only barrier. Outside of fake news, having enough money can make nearly anyone famous providing they have a unique selling point. It can make sales if aimed at the right demographic. It can also spread information - good or bad - at a particular demographic.

This isn’t to say of course that it’s all doom and gloom. My own company offers Social Media Advertising optimisation and a look into the Artificial Intelligence itself. We’ve helped companies find very particular targets in very particular niches.

Notably, tracking down people who are 50+ years old and who are interested in Bollywood and Fitness. We gained a tonne of leads, and frankly, I wasn’t even sure anyone would be interested. But it goes to show how amazing the targeting system and Artificial Intelligence system that Facebook uses actually is.

I think at this point, I’ll end the article here. This has been a fairly heavy write up and probably quite fear-inducing.

I would like to thank you for taking the time to read this article. If you’re within the marketing field, I heavily suggest you share this with your friends. If not - cool! Thanks for reading!