A shocking 2024 cybersecurity study revealed that 51% of internet traffic is now bot-generated, not human. That's right. More than half of all clicks, posts, and engagement online comes from AI and bots, not real people. Both Reddit co-founder Alexis Ohanian and OpenAI's Sam Altman have raised alarms about this phenomenon.
What We Cover:
The 2024 Imperva study showing 51% of internet traffic is bots The viral ""Pound Cake"" story: An obese cat's weight loss journey that devastated thousands of followers - only to be revealed as completely AI-generated Why Sam Altman (creator of ChatGPT) is concerned about AI-generated content taking over How AI bots farm engagement on social media platforms Swiss researchers confirm people can't tell the difference between human and AI content anymore 46% of young Australians get their news primarily from social media. What if it's all fake? The vicious cycle: AI bots creating content, other AI bots reposting it, with no humans involved Cyber attackers using AI tools to deploy malicious bots at unprecedented scale
The Scary Reality: Social media has shifted from connecting humans to simply consuming AI-generated content designed to trigger dopamine hits. As one study from the University of Melbourne found: we're witnessing a cycle of artificial engagement that no longer involves humans.
There's a vicious cycle of artificial engagement, noting that AI bots can create artificially
generated images that are boosted and reposted by other AI-powered accounts, and that really
creates a cycle that no longer involves humans at all.
And it can grow very fast, right?
So it can become very engaging and very stimulating and scary at the same time.
All right, it's another amazing day, and we're planning an amazing episode for you.
So Steve, let's talk about the dead internet theory, and it's blown up on Reddit.
Yes.
What do you know about this?
Well, so yeah, Silicon Valley, Front Row, that's our channel.
Yeah, so, you know, it's pretty amazing, David.
I don't even know, I mean, basically more traffic on the internet is bot traffic than
humans.
And, you know, once in a while you look at some articles and it says, oh, this was AI
generated, right?
Yeah.
But I thought it's a once in a while thing.
It's not once in a while.
In 2024, or almost in 2026, so this is like a year ago, it's already 51% is bot traffic,
AI traffic on the internet.
That's crazy.
That's really amazing.
So we'll try to kind of go through what that means.
So what you're saying is basically one out of two clicks or traffics on website is just
fake?
It's generated, yeah, the content is generated by AI.
I hope people that watch SVFR, I hope it's more than that for us, I hope it's real people.
Or if you question if we're AI generated or not, you can call us, we'll give you our number.
We are real people, we are not AI.
So in the future, I guess this would be a big deal.
All right, let's look at this case and try to understand it.
So the first thing that comes to what's significant is that both Reddit co-founder Alexis Ohanian
and Sam Altman, OpenAI founder or whatever, he's president, they're both talking about
the dead internet theory.
And this relates exactly to what we're talking about, to have more bots taking over.
So if in 2024 it was 51%, obviously the number is going up.
With the prevalence of OpenAI, chat GPT.
Yeah, and others.
So anyway, so these two guys are raising, and it's funny, you're exactly right, Sam Altman
is the co-founder of OpenAI who creates all of this stuff or indirectly creates all of
this stuff, but he's concerned about it, right?
Now, there was a study done by a cybersecurity company in Perva, and the study is right there
2025 study done for 2024 data.
And this is what brought up this 51% bots.
So Steve, due to a rise of AI and large language models, so these agents can rapidly create
posts alongside AI generated images, and it's designed to farm engagement.
So what this means is getting clicks, likes, comments, and to build up the reputation on
different platforms on social media, like Facebook, Instagram, TikTok.
And this is one of the contributors of why there's so many bots, because there's a lot
of fame and potential to just grow an account.
That's correct, that's correct.
Cause traffic.
You know, there's a famous saying that started with the Watergate scandal when the two reporters
were investigating the burglars from Watergate in 1972 or 74.
What did they say?
And they said, follow the money.
If you want to figure out what happens, how the things are situated, follow the money.
So that's right here in all these things, there's more money to be made through content.
And doing content through AI, it's a lot easier and it can be more amazing or more stimulating.
So it brings a lot more money.
We'll talk about this example with a cat.
One more thing is before experts dismissed this idea as a conspiracy theory, the fact
that there's more bots or the dead of the internet theory.
But now the study brought it and clearly showed that even in 2024, it's already the majority
of the traffic is bot related.
Crazy.
And so here's the story on the cat.
So this one was pretty funny.
We're kind of laughing pretty hard, but I don't know if you want to see it's funny.
So there's this really obese cat, really big cat.
This cat was adopted by someone who felt bad for the cat.
So he put the cat on a diet.
So he was kind of monitoring how the cat is losing weight and on the internet.
And he was putting posts and then through this whole weight loss journey.
And then it had thousands, maybe more than thousands of fans were watching the cat's
progress of losing weight.
What's the cat's name?
Poundcake.
Interesting name.
Yeah, maybe she eats a pound of it.
Wow.
Of cake.
Just kidding.
So what happened is somehow it was discovered that this whole thing is AI generated.
It wasn't, there was no cat, there was no person doing it.
There was nothing.
You know, it was all AI generated.
So this photo is, we don't know if it's even real.
Yeah.
It doesn't seem like it's real.
It doesn't seem like the whole story is real.
So, and also people are very frustrated because again, you can ask the question about what
does it matter?
If it's fun, it makes you happy or you enjoy it, why does it matter if it's AI generated
or not?
Well, it looks like it does.
When people found out they were very upset, they were devastated and some were crying.
Apparently the cat died.
In the story.
Yes.
And all of the followers, people that are following the journey, from the comments they
said that they were devastated, they were crying.
And then they found out, like you said, so it's like a mix of emotions.
Right.
So anyway, so that's one of the stories that actually Ohanian, the co-founder of Reddit,
mentioned in his post on X, he says that, yes, can this be the dead of the internet
theory?
You know, does this contribute to the dead of the internet theory?
Oh, yeah.
So one more element on this is that a team of Swiss researchers are saying that the internet
users now have a harder time differentiating between human created and AI generated content,
citing the rise of deep fake videos.
Now, this is kind of obvious.
You know, I don't even know if we need researchers to do this, but it seems obvious because we
look at videos and we can't tell, right?
So maybe these researchers, they had a grant they had to use up.
So they just came up with this explanation.
But that's obvious.
But the implications are pretty, pretty big.
Like case in point, Tillie Norwood, we did an episode about or one or two.
Yeah.
Hollywood actress, AI generated.
So more on that.
So this survey found that about half of US adults are more than more than half the US
adults say that at least sometimes get news from social media.
And we've seen that in social media, a lot of the material is it could be fake.
That's right.
AI generated.
Now, this number in Australia is even more concerning because 46% of people between the
ages of 18 to 24 said that social media is there was their main source of news in 2023.
So if something like this was AI generated, what if the news articles were also just conjured
up?
That's right.
And sometimes you would see under a story, under a post, under some news, you would say
this was AI generated.
Right.
But what if that somehow it's missed?
Somehow it doesn't catch it.
The algorithm doesn't catch that.
Because as much as the algorithm gets better to catch those fake stories or fake ideas,
so do the people who generate them, the program that generate them.
So it's a catch 22 and you may not actually know if something is real or not.
Yeah.
I would like the label to be on everything that's AI generated just so I know what's
real or what's not.
Yes.
One other point mentioned here by researchers is that social media has become less about
connecting humans to other people, but about consuming content and getting hooked by deliberately
targeting dopamine hits in our brains.
Also citing how generative AI used by retail companies to sell products.
So again, it's about making a profit and it's hard to resist when it comes to making a profit.
That reminds me of the early days of Facebook and Instagram when it's meant for sharing
photos and connecting with families.
But now there's ads and things are just more addictive on it.
We know about the term doom scrolling.
Once you start, it's kind of hard to stop.
That's exactly.
Yes.
So it's like consuming content, getting hooked on these things.
So I think that's a very, very crucial thing to look out for.
And you know, as much as I think I'm familiar with this and understand how this works, I
get stuck with the phone and do exactly that, keep scrolling.
And sometimes for a long time, because the content is very interesting, it's very planned
in a way that makes you kind of want to continue.
Two more points on this one.
I think they're very significant.
A study from University of Melbourne in Australia is about how there's a vicious cycle of artificial
engagement, noting that AI bots can create artificially generated images that are boosted
and reposted by other AI powered accounts.
And that really creates a cycle that no longer involves humans at all.
And it can grow very fast.
Right.
So it can become very engaging and very stimulating and scary at the same time.
How many times have you seen these comments on like your video?
You use Twitter, right?
Yeah.
Do you get messages?
And it's just something like, oh, yeah.
Yeah.
Hi, how are you doing?
Oh, yeah.
I get to know you.
Yes.
I get so many of those.
Yeah.
And those are, I think, obviously from bots.
Sometimes maybe they're people that are just trying to do phishing or scam.
And I'm going to read just one last thing that I think it's very significant because
we haven't really talked about this.
Just now, this report also mentioned how the rise in accessible AI tools significantly
lowered the barrier to entry for cyber attackers, enabling them to create and deploy malicious
bots at scale.
So this is probably the scariest thing of all, because it's a cat that's AI generated.
It's not that big of a deal, but when it comes to the cyber attackers, the ones who are going
to create havoc to your network, to your website, take down websites or systems entirely, it's
just becoming a lot easier for them to do this with these tools.
So I don't know if that may be a separate report on it that we want to look in the future,
but this seems like a pretty big concern.
Even before AI was popular, like several years ago, I know companies were being attacked
by these barrage of bots already.
Without AI, it's just bots attacking a website, like they just kind of go in there, log in
or whichever, and it slows down the service and eventually the website gets shut down
temporarily.
So you multiply that by a huge factor when AI is involved that can create this on a much
larger scale and much more damage.
Comments.
Let's look at comments from previous episodes.
So the first comment that we want to read is from GibusMoon3615, and this was on our
first episode that we made for the podcast.
That's right.
Thanks again.
So this person said, I'm not a fan of AI, unless you count simple robots like thermostats
and fridges.
I do find GPS very helpful because before I finally caved and got a smartphone, I got
lost all the time.
I'm just one of those people who gets lost easily.
However, I do try to memorize the directions whenever possible to exercise my brain, just
like David Lam said.
Oh, nice.
Yeah.
That's good.
So she realizes that you do actually use your brain more when you try to find something,
to memorize or to find a location versus using the GPS.
But clearly the GPS is useful, can be used.
We're not saying not to use it.
I think what we said in that episode was try not to use it one time, especially a place
that you somewhat know fairly well.
Right.
And you could explore and see new horizons, new streets.
Yes.
Right.
We thank the audience for the questions.
Please keep them coming, questions and comments.
We'd like to address them.
I know there have been some other ones that we haven't talked about, but in the future
we'll address as many as we can.
David, do you have any questions for the audience?
Yeah.
So question for the audience is what was one thing that surprised you on the Internet when
you realized that it was AI generated when you thought it was real?
Yes.
Give us an example on that.
We will definitely go through those in our next episodes.
And if it's as funny as this one, I'm not sure if people think this is funny, but we'd
love to hear more funny ones.
And are you worried about AI generated content?
Are you worried about the fact that the majority of traffic on the Internet now is AI generated?
Yes or no?