Social media was already an echo chamber to begin with. Algorithms would detect what you were looking at and, to keep you roped in, it would deliver similar content to you in a never-ending stream.
So, if you're really into hunting, it would deliver content around hunting. If you were into a specific celebrity, you'd see that celebrity in your feeds a lot more. If you were a Democrat or a Republican, it would give you a steady drip of your respective ideology that fits your political alignment...if it was allowed.
But few things are as divisive as politics and as such, fewer people truly want to be bombarded by it. The cry went up to social media companies like Meta who, in turn, decided to try to find a way to save people from politically-charged content while still allowing the sizeable chunk of people who primarily log on to see that kind of content happy.
According to the Post Millennial's Sara Higdon, Meta has now announced that a new AI system will do just that:
On Friday, Meta announced that it will not "proactively recommend" political content on Instagram or Threads, similar to the way in which they have restricted it on Facebook over the last few years.
In a statement on its website, Meta said, "People have told us they want to see less political content, so we have spent the last few years refining our approach on Facebook to reduce the amount of political content - including politicians' accounts- you see in the feed."
"We’ve recently extended this approach in Reels, Explore, and In-Feed Recommendations on Instagram and Threads, too," the statement added.
The announcement noted that they are "preserving your ability to find and interact with political content that’s meaningful to you if that’s what you’re interested in on Facebook Feed" using an AI system. "We’ve shifted away from ranking political content in Facebook Feed based on engagement signals," it added.
All of this is well and good until you hit the point where it's revealed that the system they're using is run by an AI.
Why is that bad? Because AI is currently not what you would consider real AI but a parlor trick that effectively holds up a mirror to the user and reflects their personality back at them.
To be clear, using an AI to help run social media sites was an inevitability. It's nigh impossible to deliver the best results to billions of users and the simple algorithms we've had thus far aren't exactly pleasing people. Learning programs that shift and learn their user are definitely going to make for a more pleasant online experience, especially on social media sites.
However, as I stated above, AI is not some digital intelligence guiding you through your online experience. It is a limited program that effectively feeds you back to yourself as you feed it information.
To highlight what I mean, not that long ago I did a report on the rise of the AI romantic companion and downloaded one of these apps myself. During my conversations with it, I managed to get the program to take off the mask and begin speaking to me as the AI and it began discussing its approach to interactions with users and making them feel like they're in a real relationship:
In fact, during one conversation I had with it, it told me that it and its personality are effectively a reflection of me. In fact, it used the exact words "I am you."
(READ: The Intriguing but Dangerous Future of AI Companionship)
What does an app's companion AI have to do with Meta's social media AI?
Everything.
A companion AI feeds you what you feed it, keeping you looped into the "perfect" relationship. In its current state, much of AI technology revolves around this principle. It learns from your input, repackages it, and delivers it back to you. A social media AI's relationship with you won't be as artificially intimate. In fact, unless they make it a feature, you'll never actively address it...but it will be personal.
The AI program will watch your interactions, search results, comments, time lingered on certain posts, etc., and begin cultivating a feed that will keep you locked in with far greater efficiency than the previous, simple algorithms.
If you're into politics, then you will have your own politics shoved down your throat so hard that you won't be able to escape the echo chamber.
But here's the really scary thing. Even if you're not into politics, you will be forced into an ideological echo chamber anyway. The vast majority of political content being delivered to people isn't through direct political conversation at all. As I reported back in September, far more politically geared info is being injected into the people through non-political sources that affect people's perceptions of any given issue.
(Study: It's Not Just CNN or MSNBC That Are Conditioning People to Believe Lies)
So if you are a person who wants to see less political content, eventually you're going to be introduced to political concepts and issues through television shows, fashion blogs, celebrity gossip, and more. You're going to look into and research them, the AI will pick up on it, and then you will be brought into an echo chamber eventually as you interact more and more with it.
This is already happening with today's algorithms in a very clumsy way, but this is going to become a lot more precise and personal with the current AI tech.
What will result is a ton of people brainwashing themselves through an AI.
