Hardly any topic was as relevant in 2023 as AI – artificial intelligence. ChatGPT, the AI chatbot from OpenAI, has started a ball rolling that will not lose momentum to this day or in the foreseeable future.
But what happens when we lose control of it? What if in the future we will no longer be able to distinguish whether we are talking to an AI or a human on the internet?
Today we’re looking at an actually absurd internet conspiracy theory that somehow sounds incredibly plausible: Are there more bots than people on the internet?
Before you continue reading: In this article we cover the
Dead Internet Theory – a conspiracy theory or a conspiracy myth. It is of course nonsense, absurd and should not be taken seriously. However, it raises a few interesting questions that are highly relevant today. These are the questions we are talking about today.
Die »Dead Internet Theory«
The dead internet theory is actually just as absurd as the flat earth theory. It was created about three years ago in a forum called “Agora Road’s Macintosh Cafe” and was described in detail by user “IlluminatiPirate.”
What’s so absurd about that? The theory is that the internet essentially died between 2016 and 2017, and most of the “people” we see online on social media and other platforms aren’t real – they’re bots.
The author of the theory argues that (of all things) the notorious forum 4chan was significantly more active before this time. He gives other reasons, none of which are particularly meaningful. It’s just a conspiracy theory.
But now it’s the end of 2023 and – wait a minute – could there have been something to the theory?
The rise of AI
Imperva, a cyber security company, has published its so-called “Imperva Bad Bot Report” annually for ten years. A study that monitors Internet bot traffic to ensure greater security on the Internet.
Bad bots are those that are programmed to attack businesses and individuals, and this report aims to analyze what share they account for in global internet traffic. In May this year, the company published its 2022 report.
The 2022 result: Almost half of all internet traffic consisted of bots: 47.4 percent, to be precise. Imperva classifies 30.2 percent as malicious. Additionally, bot traffic has steadily increased over the past four years.
With the advent of AI and LLMs (Large Language Models), it is also becoming increasingly difficult to distinguish a bot from a real human. Unfortunately, the proportion of malicious bots is also increasing, and they are becoming more and more successful thanks to better AI.
But how far are we from a world where it becomes increasingly difficult to distinguish real people from bots? Not that far.
There are now even social media influencers, YouTube channels, streamers and even virtual partners, all of which are based on AI.
Aitana Lopez, for example, is an Instagram influencer who is not real but AI-generated. Nevertheless, she has more than 200,000 followers and currently earns over 10,000 euros a month.
Neuro-Sama is a VTuber, a streamer who streams with an avatar instead of a “real” facecam. The difference to other Vtubers? It is completely based on AI. She currently has almost 500,000 followers on Twitch.
And then we have Channel1.ai: The world’s first news channel where all news anchors are AI-powered and can speak any language.
Link to YouTube content
It can be expected that we are only at the beginning of this development. Experts from the Copenhagen Institute for Future Studies even predict that by 2030, 99 percent of content on the Internet will be AI-generated.
So the Internet has not died, as the “Dead Internet Theory” describes it, but in the future it could actually happen that the World Wide Web will be populated more by bots than by people.
What is being done?
We will probably have to face two major challenges in the future:
- Which content on the Internet was created by humans and which by AI?
- Who is a human and who is a bot?
We are moving towards a time where text, product reviews, videos, images, social media accounts, news and entire websites are AI-generated. What is real and what is fake? What is true and what is a lie?
There is hope: Many companies are working on methods and means, ironically also based on AI, to recognize AI-generated content.
For example, there are the first tools that can recognize texts written by AI models. Unfortunately, these are not yet 100 percent reliable. Google and Microsoft are also working on AI recognition software for images and videos.
Unfortunately, that alone probably won’t be enough. AI content and bot detection is certainly getting better – but so will AIs and bots. Media literacy will continue to become increasingly important in the future, which is why the correct use of AI and its possibilities should be learned early on.
ChatGPT, Midjourney and Co: How AI (has) changed our lives
Duy Linh Dinh: I’m a big fan of AI and what it can do! I really think it’s one of the most important innovations in tech in recent years. Not only is it extremely useful, it can be used in all areas of life – including games, for example.
In Tekken 8, an AI learns your playing style in the background and can create a “ghost” of you that other players can play against. This way you can play against bots that feel like they are real people – awesome!
Or do you still know those support chatbots that many companies use, which are ultimately only used to redirect to a real human? I think they can’t be replaced with AI soon enough.
Nevertheless, I also think it is important that the protection mechanisms evolve just as quickly as the AI itself. Imagine if game developers filled their online lobbies with AI bots that behave very realistically, but do not do so to the players reveal that they are playing against bots? How would you find that?
In other areas, such as product reviews, reliable recognition of AI texts is extremely important. Otherwise we will no longer be able to trust any reviews in the future.
Despite all these uncertainties, or perhaps because of them, I am very excited about the future.
No comment function? Due to the expected high moderation effort, we have decided to move the comments to the forum. You can comment there as usual. Click here for the forum thread.