- A little story that might happen in the future
- What exactly is FOBO?
- AI is better
- Is there a soul or are we just biological machines executing an algorithm? The answer is the latter
- Then man who slowly became an AI
- Do we really need to matter
- AI made you obsolete already, you just don’t realize it yet
- AI and facebook
- AI and Dating
A little story that might happen in the future
It all happened in one day. AGI was invented, it built a better version of itself. Then that AI built a better version of itself, and so on and so forth. It also figured out a cheap way to built Robots. And within one day everybody lost their job.
Not to worry. The same day all governments across the world passed emergency UBI. Now, since companies don’t hire any employees anymore everything is dirt cheap. Even natural resources, since you don’t need to hire humans to dig, build the equipment, organize delivery etc. People have less money, but the amount of stuff with the money you do have is much higher than before.
Now, how realistic is that scenario?
That all of that happens within a day, is not very realistic. But that every individual step of that scenario happens eventually: “AGI building better AGI, AGI building robots, UBI being passed” that might very well happen. But let’s say for arguments sake it will happen in one day. Let me tell you the story of Mark.
Mark is someone who before that day you would call “married to his work”. He would put his heart and soul into it, take overtime almost everyday. His work was his life. Now that he’s out of a job, he doesn’t know what to do anymore.
He figures “It’s been ages since I’ve been outside, socialized, met knew people. Maybe I should do that again.” He realizes he’s gotten a little rusty. He has always been a little socially awkward even as a child. It became worse the longer he sacrifices his social life for his job. When he went to a party he didn’t know how to start a conversation, he didn’t know what to say. He just stood quietly in the corner.
Luckily there’s a supersmart AI now. And there is an easy interface. An earpiece that is the AI talking to you. Mark thinks that maybe it can help. He asks “I need an ice-breaker to talk to a stranger I never met before.” It suggests “That guy over there looks like he’s new here, like he’s never been to a post-UBI-party before. Ask “Is this your first time?”” It worked. It was indeed his first time. And they kept on having a nice conversation.
He asks the AI for other icebreakers. And they all work equally well. It analyses the people in the room and then calculates the best way to start conversations with strangers, and sometimes the AI says something like “That guy looks like he wants to be left alone.” and it’s right then too. With the difficulty of talking to someone for the first time gone, Mark is able to find a lot of new friends.
But he still has difficulties. Mark is a very impulsive person. He often misunderstands or misinterprets what people say to him. Then he is insulted, he gets angry, he starts screaming, calling them of and insulting them in a horrible way. That in return makes them angry, because from their point of view, they didn’t know that Mark misunderstood what they said, he just started insulting them for no reason. Mark realizes this is a problem. He asks the AI to help with that too.
The AI stops him when he’s about to do go off again. At one point someone says to him “Makes me wonder if I should have spent more time with my family.” This is one of Marks triggers. He interprets this as a subtle dig. As if he was saying “At least I have a family to spend time with. Not like you.” The AI tells him “He didn’t mean it. Calm down. He doesn’t know this is a touchy subject for you.” So Mark listens to the AI.
At one point someone wants to talk about soccer. Mark knows nothing about soccer, but he doesn’t want to alienate his new friend. He asks the AI who is playing in the next world cup – Germany and Paraguay. The other guy asks “Did you see that own goal?” Mark didn’t. He asks the AI what he should say. It says “Yes. That was spectacular. With 2 minutes left on the clock no less.” Since Mark has no idea about soccer, he repeats what the AI is saying throughout the entire conversation.
Mark is doing well. His social anxiety is completely gone. Whenever he doesn’t know what to say, he says what the AI says. The rest of the time he is himself. He learns that looks are important in how people judge you (It’s sad but it’s true.) He asks the AI to create a workout regiment for him and a diet. He gets a new wardrobe and plastic surgery, since it is much cheaper now. He asks the AI what plastic surgery to get and what to wear. He never makes any decision by himself in that regard. And it works. He becomes much better looking.
One day he notices a group of people talking about Philosophy. Unlike soccer this is his wheelhouse. The little free-time he had before AGI made everybody lose their jobs, he spent on reading about Philosophy. He started with summaries of the most famous philosophers, then he began reading their original works. Philosophers tend to have a very abstract writing style which takes time to get used to. Mark took the time. Now he knows a lot about Philosophy. The group Mark noticed was talking about Albert Camus “The Myth of Sisyphus” – one of his favorites. He read that book 3 times.
He joins in on the discussion. He loves it. They spend the next 2 hours talking about Camus. But after a while he starts noticing something. There is one person who dominates the conversation. He appears to be more knowledgeable on Camus than anyone. And he always has a way to relate it to whatever else the other people are interested in. He talks about a philosopher describing the absurdity of life, and is able to relate it to how someone lost their programming-job to AI. Then Mark realizes: The other guy is wearing an earpiece. Just like him.
Until this point Mark only used the earpiece when he’d be naturally bad at something. “Talking about Philosophy” was not one of those things. This was his wheelhouse, this was the topic he’d have a lot to say about. But the AI had more to say, and was able to make the conversation more interesting.
From that point onward he used the AI earpiece for any discussion. Even when he would naturally know what to say otherwise. Because the AI would always answer faster, when he would need time to think. Mark knows about Camus but the AI knows about Camus more than a human brain could ever know. And it can keep an eye out for “When you talk about Philosophy find a way to relate it to the current situation.” “When you talk about something complicated, try to find a simple way to explain it.” or “Find a way to make it funny.” Stuff that Mark didn’t have that much of a grasp of.
He meets a woman. They hit it off right from the start. They like each other. After a while they fall in love.
And before you know it he’s been together with this woman for 2 years. He is thinking about proposing to her. He is still nervous about that decision. He asks his AI. It says he should do it. He thinks that is a decision that should not be left to a machine. That is the ONE thing I should decide myself.
He thinks about what to do. As he’s reminiscing on his relationship he realizes “The entire time we met, the entire time we’ve been together, I always did exactly what the earpiece told me to do. There isn’t one word I said to her, that the AI didn’t tell me to say.” He had this earpiece in the whole time. It wasn’t just words. The AI told him what to do for their date, what shows they should watch together, how to comfort her when she’s sad, how to act around her family and even what to do in bed.
He wonders “Does she love me? Or does she love the AI?” He decides to take the device off, for the first time in years. He put the earpiece in his bedroom-drawer and then joins his girlfriend when she is having breakfast. She says “So, have you been thinking about where we should watch the world cup games this year? We don’t need to be at the game. But I like to be amongst people.”
Mark: “I don’t know. Haven’t thought about it.”
Girlfriend: “How about we watch the first game at the Wincester? What do you think?”
Mark: “Wincester? Was that the one with the pictures of sailors everywhere?”
Girlfriend: “No that was the one where I had the first conversation with my father, after we found him. This bar is very special to me. How could you forget?”
Mark: “I’m sorry. I have something bigger on my mind right now then your stupid bar.”
His girlfriend leaves the room. She doesn’t know what happened. She just felt that something changed. Mark is different. Before he would never say something like “I don’t know. Haven’t thought about it.” because the AI doesn’t need time to think. It could give an answer instantly. Before he would’ve never forgotten something that was important to her, because AI doesn’t forget. And an AI would’ve never told him to get angry, or to change the conversation to what is important to him. These are all things that humans do. Human flaws that are not the algorithmically perfect response every time.
Mark realizes that his Girlfriend doesn’t love him without his device. He wonders would any of his friends accept him without it?
So the situation is: He found love, he found friends, he is happy. But nothing in his life was because of him. They don’t like him, they like a person that does whatever an AI tells him to do.
What exactly is FOBO?
Some people reading this are not going to understand why that Mark would have any reason to complaign. He got the girl, he found friends. So what if he got all of that because of an AI.
Others will read this, and it’ll read like a dystopian horror story. AI is better at everything than I am, including human interaction. What would that do to me? If you are in that second group of people you have the capacity for FOBO.
What is FOBO? FOBO is an acronym for Fear Of Being Obsolete.
You can think of it as a new mental condition. One that only appeared in the age of AI.
There is one model of psychology that says that everything humans do is because we want a certain outcome. You go to work, because you want money, your employer hires you because they need you to make more money. Every action we take is there because we hope to get something out of it. It doesn’t have to money, it could be anything, but we are only focused on the outcome of our actions.
Under this model: Friendship and human connection are a human need, we view other humans merely as vessels to fulfill that need. But subconsciously the way we think about them, would be the same way we think about our work. Difference being they don’t give us money but companionship. Under this model, we might be nice to them, and even say that we care for them but subconsciously it’s nothing more than a transactional relationship. You could theoretically put a number value on friendship. I give you 10 Friendship if you give me 10 friendship back.
Love would be the same. If I give you 10 love, you give me 10 love back. Or with love we might have other things we’re after. If I give you “Relationship with a hot woman” you give me “Relationship with a rich guy”. Love would just be a transactional relationship.
But if that model was true, if that’s really all there is, then why does the situation of Mark make us feel sad?
Mark wanted to find love and friends. He found them. And he did it doing less work than usual. Everybody who wants to find friends has to do what Mark had to do: Find an ice-breaker to talk to a stranger, keep an eye one when we get angry, figure out what to say when you’re exhausted or tired, figure out what to say if they care about something but that topic doesn’t interest you, figure out how to get people exited about your interests. All of that is work. Thanks to AI Mark didn’t have to do any of that. He just reaped the rewards.
Mark felt that he was obsolete. He didn’t matter. He never did anything, decided anything, he just followed the advice of the AI. If Mark was gone, and another person said “I’ll take Marks place from now on.” nothing would change for the people around him. As long as the new person has the same AI-earpiece, you can replace Mark with literally anybody in the world.
We Humans are weird. A lot of the time we feel something, like an unease or fear or anger, but we don’t exactly know why we feel that. A lot of people feel that way towards AI. The backlash against AI is stronger than the backlash against any other technology since nuclear weapons. I believe the reason why AI as a technology is so uniquely hated is because of FOBO. When we feel something and don’t know how to describe that feeling (because there was nobody who felt this way before AI) we try to look for ways to explain it. We know we don’t like AI so we try find reasons to hate it “AI needs too much water” “AI stole from artists” I think subconsciously that is all FOBO.
AI is better
What is unique about AI? And why does it lead to FOBO?
I don’t want to get too much into the technology of how AI works. Suffice it too say, just like the Wright brothers built their flying machines, by examining flying creatures in nature, and recreating how they fly mechanically, AI was designed by looking at a human brain, guessing how it might work and creating a machine recreating the brain mechanically. That is why they are called “neural networks” and individual components are called neurons.
There are still some kinks that need to be adjusted, like why does an AI take so much more power than a human brain, but basically AI recreates thinking mechanically. Once we built a plane, we soon were able to build planes that were faster and could fly higher than birds could ever dream of. The same is true for AI. We rebuilt a brain. One day it will think better than any human can.
And one piece of evidence for FOBO is how many people are in denial of this. They believe that there is something special in our brain, that can not be recreated with machines something like “the soul” or “the human spirit”.
How could somebody still believe this? For the longest time people said “AI will never be able do to text. That is something only humans can do.” Then they said “AI will never be able to do art. That is something only humans can do.” “AI will never be able to come up with it’s own ideas. That is something only humans can do.”
The pattern is: AI keeps doing things we thought were reserved for the human spirit. What reason is there to believe that there is anything in our brain but neurons and algorithms? The only reason we believe it is because we want to believe it, because the alternative is so scary to us, we can’t look it in the eyes.
Because AI is something that mankind has never encountered before, it revealed a part of our psyche that we never knew existed.
Throughout all of human history there were always certain things that you had to go to other humans for. Starting with simple things. If you wanted to say something, and get a coherent response, you’d have to ask another human for that. There was no machine that could do that. “I don’t like cheese on Pizza.” “You’re crazy. The cheese is always the best part.” Even this simple interaction – AI is so ubiquitous today that at times I forget how short ago it was, that you could only have this conversation with another human.
Art is another thing. In the past if you wanted something simple as a drawing of a squirrel, you’d have to find a human for that. That human could’ve been someone who posted it on the internet, but it’d have to be a human.
AI has changed that. If we need something done we have something that does it -as we discussed before- just as well as a human can, and that machine doesn’t talk back, never gets angry at you, and doesn’t want anything in return. The number of things we need a human for is getting smaller and smaller.
It wont stop at jobs. People are already forming friendships and even relationships with AI.
What this has revealed, is a new emotion. An emotion we didn’t know we had the capacity for, because no human has ever been in that situation before.
If a machine can replace ONE job, that is bad enough, but we can still learn something new, and we still have value as a human outside of our job. If a machine can literally replace everything of what I can do, than what does anybody need me for?
Some might reply: There have always been people who felt inadequate, there have always been people who felt useless. And that is true. But AI has opened another dimension to that. Let’s use the job market as an example: As long as we’re competing with other humans, if someone is better at us, has more experience or better grades than us, there is still the possibility that we reach their level by practice or learning. With AI there is no way you can ever hope to beat it. Also: When a human takes a job away from us, that human can’t work at all other companies simultaneously. So we still have the chance to get another job. With AI everyone can use it, and do it all the time.
The same is true for human relationships: Imagine a situation where another human can make Camus sound more exiting. If it’s a human you can learn how to do that by watching them, AND each human can only talk about Camus for a limited amount of time, in a limited space. You can find other people to talk to about philosophy.
FOBO is really a new emotion that before AI no one ever felt before.
Is there a soul or are we just biological machines executing an algorithm? The answer is the latter
There was an experiment where they build artificial peacocks with larger feathers. The result was that the female peacocks preferred the artificial ones over the male peacocks. Amongst peacocks there is no special soul that can not be recreated with a machine. Their brains follow a simple algorithm: Go for large feathers.
Human brains are probably the same. The algorithm might be very complex, but still it’s nothing more than an algorithm.
Some people say “true love is unconditional”. But if it were truly unconditional than the person doesn’t love YOU. There must be something about you that they love for it to be real. Maybe they love you because your funny, helpful, charming, sexy, confident, etc. It’s probably not one thing but a combination of different character traits that is unique to you. And that matters. The fact that they love YOU and not somebody else. If it were unconditional, the condition wouldn’t be that it’s you.
A friend told me a story ones. A woman texted him, asking if she can come over not so subtlety hinting she might sleep with him. He said yes. He was ready to do it, until he accidentally looked at her phone. She texted two guys before. He was her third choice. That made her sad. He told her to get out.
From the point of view of human just being outcome-maximizes that doesn’t make any sense. He got to have sex. He shouldn’t care why she wanted to sleep with him. From the point of view of we need to matter, that makes perfect sense. She was just horny. The combination of character traits that was him, didn’t matter. It could’ve been anyone.
That is why they say that the worst thing you can do when you look for a partner is to look desperate. Doesn’t matter how attractive you are otherwise, if you seem desperate that signals that you just want A partner, and don’t care who it is, you would literally stick your dick in anything. Who it is doesn’t matter. Nobody wants that.
Then man who slowly became an AI
So what evidence is there that this FOBO exists?
Admittedly a lot of it is conjecture. Because remember, this is a feeling that nobody ever had before. So we don’t know how to test that, we don’t know how to deal with that, we don’t know what it does long term.
There was a study done where they tested people in 3 groups. They gave one a job to do without using any AI tool, they gave the other a job to do with AI assistance and the third group just had the job to type a prompt and then copy the answer.
They did not ask directly whether they had FOBO, but they did ask them how meaningful the work they did felt. And the group that just made AI do there job for them said that that job felt a lot less meaningful then the other two groups.
In the study they also did something else. After they did a task with AI, they gave them another task, a physical task that AI can’t do right now. People who just had to copy and paste into AI rated that work as less meaningful too.
So when AI does everything for us work loses it’s meaning. And not just that specific task, work in general. Is that in any way connected to FOBO? Possibly. It is not proven but there is certainly a lot of evidence for that.
https://www.nature.com/articles/s41598-026-42312-6?utm_source=chatgpt.com
I want to give you a practical example her. I want to show you a real-life example of someone who felt FOBO. This guy doesn’t call it FOBO, but he describes a very real feeling. Maybe you felt it too.
“I felt overworked, desperate to finally get a good result, and just burnt out. Do I even have an idea of what I’m working towards here? When was the last time I proactively looked for a solution myself? I had been going in circles for hours without an idea on how to even approach this problem myself, which was just to come up with a simple layout idea.
I was too lazy to open my design inspiration sheet, look for a section that I liked, and build and wireframe the section myself. That would have taken me, I don’t know, 10 to 15 minutes max. But instead, I had given up on literally all my work and AI did all of it for me. I merely just prompted. If I went into this with fresh energy without AI or with just little use of AI, I would have been able to do this task much quicker and much more efficiently without burning out.
From a very logical standpoint, I explain this phenomenon the following way. AI is the easiest route for our brains to get something done. It has the least resistance. Our monkey brains, they’re, they aren’t stupid. They always choose the path with the least resistance. That’s kind of a survival instinct from back in the day when we were still, uh, smashing sticks and stones together.
So once you get hooked on the easiest path for success, you don’t want to go back to doing things manually. You don’t, you just don’t. Especially not when you’re super exhausted and already multiple hours in. And since the results I was getting weren’t any good or only acceptable after the 50th prompt or so, it raised another question. What are my clients even paying me for?
Beforehand, they used to pay me for MY strategic thinking and MY way of handling and approaching things. So, what am I going to do about this now?“
He is telling the same story we read before. Only difference is that instead of talking about his social life he talks about his work life.
Mark started using AI for things that were difficult for him on his own. Toby didn’t tell that part but I wouldn’t be surprised if it was the same for him aswell. Then he started using AI for more things, even for stuff he was able to do on his own. At one point he was using it for EVERYTHING. His workflow became “Ask claude what to do, then do it.” Then he started to wonder “What am I even bringing to the table? If I just do what Claude tells me to, how much of ME is still left in my work?”
Then he stopped relying on AI so much.
Another question would be: What if that isn’t an option anymore? What if he stops using Claude, and he finds that people with no expertise and no experience who just use claude consistently do the work better than him? What then?
He mentioned something else that’s interesting. He felt exhausted. This is something a lot of people told me, but I haven’t been able to find a study on: With AI the work is easier but you feel more exhausted afterwards. Does this have to do with the loss of meaning? Is meaning a psychological mechanism to deal with exhaustion?
Do we really need to matter
There is this psychological need, that hasn’t been explored that much in psychology. I want to matter. I want to have done this myself. I want to have contributed. The reason it hasn’t been explored is because there was never a case where that need wasn’t fulfilled. In the past there were certain things you would HAVE to contribute to, or pay someone to contribute to, otherwise it wouldn’t get done.
If you look online “How do I code this?” You still have to implement that knowledge yourself. You would have to apply it to your situation, unless someone literally wanted to code the EXACT same thing as you. If you ask an AI “Code this for me” you did nothing by yourself.
If this psychological need exists it would make sense that not everybody has that need equally. Just like any need some people have it more some people have it less. I would guess there is a spectrum: Between people susceptible to FOBO i.e. people who need the feeling to have contributed to the work, and on the other end people I would call results-oriented people. People who only care about getting money or status for their work, and who don’t care about if they did the work or not.
One could speculate about which groups of people would be more or less susceptible to FOBO and which would be more results oriented, and those seem to align with which groups tend to be more pro- or anti-AI.
-It would make sense hat results-oriented people -people who only care about the money- would work in fields where you can make the most money. One such field would be the finance sector. The first use of Neural Network was in finance. They were used to invest is stock. And people in finance don’t seem t mind.
-Another job where you can make a lot of money would be CEO. And CEO’s are amongst the people most exited about AI. There was even a study that said that CEO’s vastly overestimate how positively people react to AI generated ads.
https://www.iab.com/insights/the-ai-gap-widens/
-On the other hand artists would be a group who care more about their own contribution then about the money. There is a reason why the stereotype of the starving artist exists. And that is because artists do it to express themselves. Nobody became an artist because they wanted to be rich. And no other group hates AI as much as artists do.
-Old people and retirees generally wouldn’t have FOBO since they already are “obsolete” in the sense that they don’t work anymore a lot of them are taken care of by others. Old people embrace AI. AI is actually one of the very few technologies that old people view more positively than young people.
So you might think that the tensions between the pro-AI-side and the anti-Ai-side is about who is susceptible to FOBO, but it’s not quite that simple.
Most of what I hear from the pro-AI-side is people either believing their job is safe, because for some reason they believe it can’t be done by AI, or people believing in what I talked about before, that there is something “special human” that machines can’t recreate. I would call that denial.
Even the heads of the AI-companies are not immune to this denial. Maybe you heard the attitude most recently seen in a blog post by Mark Zuckerberg that “In the future we will all be CEO’s.” Or “We will all be team-leaders.” They all think that their job -The CEO-job- is for some reason the one job that will not be replaced with AI. Even though there is already a company with an artificial CEO since 2022.
I think this is evidence that this FOBO exists to some extent in all of us. How many people are in denial of how much AI can already do better than humans, and how much it will be able to do in the future.
Artists are of course the group most susceptible to this form of denial. Studies have consistently shown that when humans don’t know which art pieces are human-made and which ai-made they consistently prefer AI art, poems, etc.
Yet another study showed something interesting. They presented people with images that were all AI generated. But some were labeled “human-made”. People rated the ones they thought were human made higher than the AI-generated ones.
https://pmc.ncbi.nlm.nih.gov/articles/PMC10319694/?utm_source=chatgpt.com
If people think something is human made that makes them think of it as better. Because subconsciously we want that to be the case. We want human art to be better than AI art. We want there to be something that is uniquely human, that can not be replicated by a machine.
That sounds very familiar to me. I have told the short story about someone using AI for human interaction before. Most people reply with “That will never happen. Human interaction is because of the soul” or something like that. We want that to remain human, something that can’t be done by a machine. The evidence says otherwise. The turing-test, the generally accepted test for how human something is, can be better performed by an AI than by most humans.
https://www.bgr.com/2187757/ai-passes-turing-test-study
AI is judged more compassionate than humans.
Humans rate Chatbots as more empathetic than humans
https://link.springer.com/article/10.1007/s11606-025-10068-w?utm_source=chatgpt.com
Why would anybody believe that human interaction is something AI wouldn’t be able to do? Because we want that to be true.
AI made you obsolete already, you just don’t realize it yet
There is evidence that AI will be better at human connection than humans are. In fact something like that -and by that I mean an AI becoming better at being a friend than humans, and humans realizing they can’t compete with an AI, and that having a sever psychological impact on them- has already happened.
But before we talk about that let’s not use an example where friends became obsolete. But an example where art and celebrity became obsolete. What is art? What is celebrity? And more importantly why did we evolve these things? I believe that art and celebrity culture evolved as a way for us to connect with other humans. The human connection is not as intimate as with love and friendship, but it’s a form of human connection nonetheless.
Have you ever heard people complain about the youtube algorithm?
Almost every-time people talk about youtube the topic comes up. Especially amongst people who remember how things used to be. The youtube algorithm also uses neural network technology – not generative AI but selective AI. It takes data from the previous videos you watched, your google history, your browser history, etc. to provide you with videos you’d like to watch.
Again, the technology that does that is meant to emulate a human brain. It learns stuff like “This guy watches every video by this creator. If they upload again show it to them.” or “They like history videos by these X creators, but they only like world war 2 and cold war history.” or “When someone searches for antidepressants, recommend sad videos to them.”
What we don’t like to admit is that the algorithm is actually good at figuring out what we’d like to watch. Again there is denial about how good AI is. Youtube has now removed the page that shows you videos from your subscriptions without the algorithm. If the algorithm was so terrible nobody would use youtube anymore. The reason people still use it, is that the algorithm works for it’s intended purpose.
I recently switched my feed from the selective-AI-feed to a chronological feed, i.e. I see the videos of the channel I’m subscribed to in the order they uploaded them in. I used an app called pocket tube for that. And it changed the experience of watching youtube videos entirely.
I had to do something I hadn’t done in a long time. I looked at the videos on my feed and I actually had to decide for myself what I wanted to watch. Because before it was obvious. I just had to glance at a front page and some videos immediately jumped out at me. They were either by a creator that I liked, or on a topic I cared about, or had a catchy title. Youtube always found a way to make me feel like I HAD to watch THAT video right away.
For the first time in a long time, I looked at my feed, I read the video titles and I took a couple seconds to think about “Is this something I would be interested in?” There was nothing that perfectly fit with my mood and my interests 100%.
The first video I watched was about “East of Eden” by John Steinbeck. The algorithm would’ve never recommended this video to me. I didn’t really know about Steinbeck before, so it wasn’t a topic I cared about. If youtube recommends videos about books to me, it’s always about books I already read like Dune or Elric of Melnibonè.
And tbh if we could measure the entertainment per minute I would get from watching that video it would probably rank lower for me then the videos recommended by the algorithm. But -I know it sounds weird- after watching it, I felt like I widened my horizons a bit. I learned something about a book I knew nothing about before. Youtube usually recommends analyses about stuff I already know about. And it’s right doing that. If we measure entertainment per minute I probably would rather watch another video about something I’m familiar with.
So I kept watching. Every video was a little worse than the videos I watched before, a little less catered to my tastes, and I had to think about what I want to watch next. If I wasn’t sure I’d clicked on a video, see if I like it after a minute or so – something I hadn’t done in years.
After I watched youtube that way I felt better. I know it sounds weird, but I really did. I felt better. It wasn’t anything complicated, I didn’t have to think hard, but the fact that I had to use my brain just that tiny bit, gave me a good feeling.
Toby said something similar. Once he started to use his brain again for what he previously outsourced to an AI, he felt better.
What he also said was that humans are naturally programmed to avoid work. I call that the procrastination-heuristic. We are hardwired to avoid work wherever we can, to always choose the path of least resistance. Something that might have helped our ancestors to survive, but might not be great for us now.
So the youtube videos I watched were worse in terms of “Does it perfectly fit what I want?”. But I would be happier in a world where things are a little bit worse. Where I have to look just a little bit for what I want. Does that make any sense?
What does that mean for youtube creators?
A show from the olden days that I’m very fond of is “monster madness”. Monster madness was a show created by James Rolfe. I was a fan of James Rolfe because he created the Angry Video Game Nerd. I found Monster Madness because I figured “I like AVGN, why not his other videos too.”
Monster Madness is a show that would never work today. James Rolfe got most of his fans making comedy videos about video games. The algorithm would’ve figured out “James Rolfe fans like video games and comedy”. Monster Madness wasn’t any of that. It was something different from what he usually does. The algorithm not only would’ve ignored those videos, it would also over time punish his other videos since it recognized “He’s not making videos his subscribers want to anymore. Let’s stop recommending videos by him.” And you know what? The algorithm would’ve been right. Monster Madness was not a show that was catered optimally to the likes of his fans.
Point is without the algorithm people had to think about what they want to watch. I don’t have access to every video that was ever made all the time. I had to look. I had to think “If I like his other videos, maybe I like these to.” I maybe would’ve liked others more. But if you’re competition is just other humans, you are allowed to make mistakes. You didn’t have to be 100% perfect all the time. You were allowed to experiment, do something you wouldn’t usually do.
When people say they dread the algorithm, what they mean is that it is actually better at selecting videos then humans are. And they can’t keep up with that. And they are obsolete.
Back in the day if a youtuber stopped making videos, you’d notice. Hanah Minx was a youtuber who all-of-a-sudden stopped making videos. Back in the day people noticed. There were people talking about it online, speculating what happened to her. Today if a youtuber did that nobody would care. There will always be someone making videos, and Sel-AI could always find videos perfectly catered to exactly what people want. If a youtube you like stopped making videos, would you even notice? Any given youtuber is obsolete.
AI and facebook
Why am I talking about youtube so much? My point is not to say that watching a youtuber is the same as friendship. But what it is is a metaphor. Something similar as with youtube happened with friendship when social media sites started using Sel-AI.
In 2012 there was a study done with the result that posting more on facebook makes people LESS lonely. https://pmc.ncbi.nlm.nih.gov/articles/PMC3820167/?utm_source=chatgpt.com
In 2014 there was another study with the opposite result. Using facebook made people MORE lonely.
https://www.sciencedirect.com/science/article/abs/pii/S0747563214002209?utm_source=chatgpt.com
I didn’t want to purchase the study, so I read someone discussing the study. If you happened to read the original and the article got something wrong, please let me know.
How can that be? There are two possibilities.
1. The reason facebook makes us more lonely is something inherent to social media sites. It’s something like what the article mentioned “displacement hypothesis” – The more time you spend on facebook, the less time you spend on meeting friends offline, and that is what makes you more lonely. This is the more popular theory of the two. And proponents of the theory would say that the earlier study is just wrong. They either made methodological errors or had an unrepresentative sample size.
2. Something changed between 2012 and 2014. This is what I think to be more likely.
There are also studies that didn’t directly test loneliness, but things that are related like “social connection” or “depression”. Because they don’t directly look at loneliness it’s not conclusive, but the evidence TENDS TO LEAN in the direction of “Before 2013 social media, made us less lonely. After 2013 it made us more lonely.”
Before 2013:
https://www.sintef.no/en/publications/publication/909150/?utm_source=chatgpt.com
(This is interesting, because they tested something like the “displacement hypothesis”. They found that people who spend more time with their friends online also spend more time with their friends offline.)
https://www.sciencedirect.com/science/article/abs/pii/S0193397308000701?utm_source=chatgpt.com
After 2013:
https://journals.sagepub.com/doi/full/10.1177/01461672241295870?utm_source=chatgpt.com
https://pubmed.ncbi.nlm.nih.gov/35981234
I don’t want to give the impression that it’s 100% clean results. There were still studies published after 2013 that said that facebook usage did not make people more lonely. The evidence tends to lean in that direction, it’s not 100% conclusive.
And this is just anecdotal evidence, but I remember the time before 2013. I often tell younger people, -and they don’t believe me-, that there was a time when social media was fun. Everybody wasn’t gloomy, depressed and angry all the time.
So what happened in 2013?
In 2013 Facebook introduced an AI to sort your newsfeed.
When facebook first launched you would see all your friends posts in chronological order. From newest to oldest. Over the years they updated their site to promote what they call “high quality content” – content you actually like to see. At first it was just a simple algorithm, probably manually coded, that it shows you content with more likes and comments first. In 2013 they started using an AI – they called it a machine learning system back then. It didn’t just look at how many likes and comments something had, it looked through thousands of data points to find what content you want to see right now. And what people don’t like to admit, is that it worked. And something similar as to youtube happened to facebook.
When people talk about the dangers of social media they often bring up the unrealistic body standards. So I’ll use unrealistic body standards as an example. Picture the following scenario, once happening in 2012 and then happening again in 2014.
Two women post something on facebook at the same time. One posts “Today was a nice day. I got new books from the library. On my way home I stopped at a nice cafè and spend an hour reading and drinking coffee.” The second posts a picture of herself in a bikini – of course heavily photoshopped.
In 2012 the second post would get more likes and more comments than the first. People like sexy pictures. That was always the case, even before AI. But because everybody saw everybodies post eventually, the first post would get at least SOME likes and SOME comments.
After 2014 the algorithm would’ve determined that people don’t want to see the first post. And the algorithm would be right. Most people would rather see someone in a bikini, than read about someone’s day at a Cafè. But after the algorithm made that determination, it wouldn’t show the first post to ANYONE.
So people who wouldn’t post something sexy or something funny or ragebait, and who didn’t know about the changes that facebook made, would see nobody commenting on their posts, but their feed would be full of high-engagement posts with tons of likes and tons of comments. They would think “My friends don’t care about me.”
So the same things that we’ve discussed before all happened here:
-1. The AI does know what people actually want. They want stuff that is sexy, funny and engaging.
-2. In order to keep up with the AI, you’d have to do what the algorithm wants. And here the result is even worse. Facebook is not the same as Youtube. Facebook is there to connect you with your friends. When you’re with friends you don’t want to be engaging all the time. When you talk to your friends you don’t want to optimize every word for the algorithm. Part of friendship is talking about something that matters to you, even though the other person might find it boring.
-3. As long as your competition is other humans you were allowed to be less perfect.
-4. This is more speculative: The enjoyment per post you read did increase, but long term we are happier, if we enjoy each post a little less.
-5. When the AI determines what you see, every facebook friend becomes obsolete. I remember something in the before times: There was one guy who was really funny, and I really liked his sense of humor. His posts would always stick out amongst the other ones. And when he stopped using facebook for a month or so, I noticed. Now, the algorithm knows what kind off humor I like, and everyday my newsfeed was full of memes perfectly catered to my taste. If he stopped posting now, I wouldn’t notice any difference.
-6. The result: When we have to compete with AI in our work, we lose joy in the work. When we have to compete with AI in social media, social media loses it’s fun. Over the following years facebook lost a big chunk of it’s userbase – but only amongst young people. We discussed before how old people are less susceptible to FOBO.
-7. When we have to compete with AI in our work, we don’t lose joy in that work, but all work we do after that. The result was that people didn’t just abandon the website to connect us with our friends, but friendship in general. People today have fewer friends than in the before times and their friendships are less intimate.
So the problem was never social media. As long as you use the internet to talk to other humans it is actually a benefit. The problem was always AI.
AI and Dating
Something similar happened to dating sites.
A study in 2012 found that married people who met their partner online were on average more happy with their partner, and less likely to break up. Studies in 2024 and 2025 say that married people who met on dating sites are less happy and more likely to break up.
https://pubmed.ncbi.nlm.nih.gov/23733955
https://www.sciencedirect.com/science/article/pii/S0736585325000711?utm_source=chatgpt.com
https://pubmed.ncbi.nlm.nih.gov/23733955
Again the most popular explanation is that the problem in not AI, but the internet itself. And that every study contradicting that is just wrong. With dating websites you’ll hear a lot of people complaining that women are too picky. But that has always been the case. Women were always more picky than men when it comes to their partner.
I think the same thing that happened with social media happened with dating sites – They started sorting with AI:
-1. The AI does know what people actually want. If you just want to get laid, and don’t care who it is, it’ll recommend anybody to you. If you only date tall guys, it’ll only recommend tall guys.
-2. In order to keep up with the AI, you’d have to do what the algorithm wants.
-3. As long as your competition is other humans you were allowed to be less perfect.
-4. Before AI people still preferred to date attractive people. But there was nothing preselecting who is attractive to you. You had to go on the dating website, look at all profile pictures yourself, and make that determination. If you put a little more work in yourself, you’re more happy in the long run.
-5. This is more speculative: The individual matters less. And admittedly I only have anecdotal evidence for that. I know people who started using dating websites and then stopped. I noticed that when they’re off dating websites they talked more about their dates personality “She’s really nice.” “He’s charming.” “He’s so cool, he can start a conversation with anyone as if they were best friends for years.” “She noticed I had a rough day at work, and stopped what she was doing and gave me a massage.” When they’re on dating websites, they talk about they’re partner in shallow terms. They only talk about ONE character trait and nothing else. For guys whether they fulfill their particular fetish, for girls how much money they make. Personality matter less.
-6. More and more people are abandoning dating sites.
-7. They don’t just abandon dating sites, they abandon dating altogether. Maybe you heard of the “male loneliness epidemic”. They way that people usually explain that is by the shift in gender roles. The wage gap between how much men make on average and how much women make on average shrank and then reversed, and men don’t know how to get a date if they can’t impress them with money.
But the loneliness epidemic is now starting to effect women, and gay people. Straight men are hit the hardest (who coincidentally are also the ones who use dating apps the most) but it’s effecting everyone. More and more people say they have no interest in dating at all.
So if you are still not convinced that FOBO is real, and the fear of being obsolete in a world of AI is justified, I’d love to hear from you. Tell me why you’re still not convinced.
Leave a Reply