Digital Déjà Vu

I was scrolling YouTube last week when the following video caught my eye:

As soon as I saw it I thought, “Wait, haven’t I seen this before?” I searched “impossible curl” on YouTube and I was right.

Will Tennyson had done a similar video with the same title nearly two years prior:

But here’s where it got weird. In the search results, there was another video by the channel “Browney” with the exact same title as Tennyson’s but published 15 months after Tennyson’s original video:

I know that content creators tend to post about similar topics, but this was too alike. What the hell is going on?

I decided to dig deeper by looking through some financial videos and found the same thing. For example, Mark Tilbury uploaded a video titled “How I’d Make $10k a Month If I Had To Start Again” on January 8th, 2024:

Two days later, on January 10, 2024, a video titled “7 Realistic Side Hustles to Make $1000 Per Month in 2024” was released by Humphrey Yang:

Notice any similarities?

What about this video titled, “6 Steps That Made Me a Millionaire Before 30” which also coincidentally dropped on January 10, 2024?

While the content in each of these videos varies, the way they are being pitched to scrolling eyeballs is nearly identical. And I am not being farfetched when I say that there are countless examples of this. Look at this one:

The top video was published on the February 4th, 2023 and the bottom one less than three weeks later.

And then, six months after that, Ali Abdaal published a video on a similar topic with a very similar thumbnail:

While I am not here to pick on any of these creators individually, it’s obvious that this isn’t a fluke. It’s intentional.

I did more research and discovered that what I was seeing was an art form of its own. It’s known as YouTube thumbnail design and it has a singular purpose—to drive more clicks. Here’s how it works:

1. You pay a designer to create some compelling thumbnails about a video you recorded.
2. You A/B test those thumbnails to see how many clicks each one receives.
3. Once you have the best performing thumbnail, you make that thumbnail the permanent one for your video.
4. Rinse and repeat.

The only problem with this process is that’s it cumbersome and can be quite expensive. Some YouTube thumbnail designers charge hundreds of dollars per thumbnail for their work. And while you probably wouldn’t test hundreds of thumbnails, you can see how quickly this could add up.

Of course, that’s usually how it works. But, there’s a much easier way to go about this whole process—steal.

That’s right. Instead of doing all the A/B testing yourself, just find what worked and copy it. Copy the concept. Copy the title. Copy the thumbnail. That’s what many of the creators above did, and it worked. They got their views and they grew their audience.

But, guess who is paying the price? We are. We are being A/B tested into a world of commoditized content. Things look the same and sound the same because they basically are the same. It’s digital déjà vu on a massive scale and I’m afraid it’s only going to get worse. As Kyle Chayka wrote in Filterworld: How Algorithms Flattened Culture:

Filterworld culture is ultimately homogenous, marked by a pervasive sense of sameness even when its artifacts aren’t literally the same. It perpetuates itself to the point of boredom.

This is what I’ve been seeing on YouTube and throughout social media in general. Similar conversations about similar topics with similar hooks. It’s the nonstop Twitter threads. It’s the oversharing on LinkedIn. It’s anything and everything that explicitly tries to maximize engagement.

My own content hasn’t been immune to this either. Last year I went down a SEO rabbit hole and almost quit writing as a result. I felt myself creating the same commoditized content that I was so vehemently against. 

And, this strategy worked, but at what cost? What good is creating content if you don’t control what you create? 

And that’s ultimately what this is about—control. It’s what we give up when all the content is meant to maximize a single metric. As a result, we don’t control what we see and many content creators don’t really control what they produce. They just follow what works. And in doing so, these systems come to control us.

I recently came to this realization after this TikTok video was published showing how men and women can be shown very different comments under the same TikTok video. The TikTok video in question is about a woman who is waiting on her boyfriend to come home, but takes a long time before he gets there.

When the video is viewed by women on TikTok, the top comments are generally supportive of the girlfriend and negative toward the boyfriend. However, when that same video is viewed by men, the top comments are all about how the girlfriend is being too needy and should get a life outside of her boyfriend.

Here’s the crazy part though—each side doesn’t see the others’ comments. Even when they keep scrolling down, women don’t see the pro-boyfriend comments. The same goes for men and the pro-girlfriend comments. They don’t see both sides, just their side. It’s like each user is experiencing the crowd’s reaction in their own isolated world.

Of course, this isn’t done explicitly based on the sex of the user, but based on how they behave in the app. The algorithm is deciding based on their behavior (as a woman or a man) to show the comments that would lead to the most engagement. Unfortunately, outrage and division are more likely to drive engagement than a balanced opinion.

Now think about this idea on a larger scale and a lot of things that seem puzzling start to make sense. Why do so many people feel like the economy is doing badly though the data looks quite good? Why is the vibecession, as Kyla Scanlon calls it, a thing?

It’s the content monoculture at work. The algorithms influence what gets watched, which influences what kind of content gets produced, which influences what we see, which influences what we believe, and the cycle repeats. While there’s always been negativity bias in the media, the system has never been more efficient at distributing that negativity than today.

Just look at how politically polarized the U.S. has become in the last few decades:

You don’t think that a place where anyone can express their opinion (anonymously) about anything they want would lead to this? Of course it would.

And guess what? The same thing is happening in the personal finance space. I promise you that all those stories about young millionaires and overnight crypto fortunes are not meant to make you feel good. They are meant to create anxiety.

It’s no wonder why we’ve seen a rise in financial nihilism among young people. When the mainstream media convinces you that you are doing badly financially, you may come to believe it. Of course, there are some people that are truly struggling, but there are far more who are influenced to think they are.

You may think that our content ecosystem has nothing to do with your finances, but you’d be wrong. Because it’s all interconnected. What content you consume influences your thoughts and actions. It influences how you treat others. It influences what financial decisions you make and much more.

How many people got sucked into crypto because they heard about a random altcoin that went up 10,000% overnight? How many people bought Gamestop because it became an internet meme? 

I don’t know, but I do know social media has had a huge influence in both of these cases. Does this surprise you though? It doesn’t surprise me. It actually reminds me of Bo Burnam’s Welcome to the Internet, where he states:

And it did all the things
We designed it to do

We’re living in the world of our own making and it’s getting harder to escape. I say this as someone who uses social media professionally (to share financial content) and even I’m having difficulty limiting my use. My only advice is to be careful what you consume, because it may just consume you.

Thank you for reading.

If you liked this post, consider signing up for my newsletter.

This is post 406. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data