There is a theory abroad that the current Internet is “dead” and most of the users active on the Internet are robots. They produce a large amount of content and also guide public opinion.
This theory is called “Dead Internet theory” is highly discussed abroad, but it seems that no one is discussing it in China.
Advertisement
The theory has two main components:One part is that robots have replaced human activities on the Internet; the other part is that actors are using these robots to manipulate netizens.
When the activities of robots on the Internet exceed that of humans, some people think that this means that the Internet is “dead”. In this theory, the “death” date of the Internet is between 2016 and 2017.
The main basis for this date is that in 2016, the security company Imperva (one of the leaders in network security) released a report on robot access traffic, revealing that robots accounted for 52% of the Internet's network traffic.
Advertisement
It is difficult to verify where the dead internet theory was first proposed. It is often regarded as just a conspiracy theory, but there may be more to it than that.
Internet Robot Content
In fact, today’s robots not only use more than 50% of network traffic, but also produce more than 50% of the content. Especially on social platforms, robots not only produce articles, videos, pictures and other content forms, but even comment areas Much of its content is also produced by robots.
You will find that many of the existing social platforms on the Chinese Internet have introduced official comment robots. Some will directly tell you that it is an AI comment, while others will not tell you.
Picture: This is Weibo’s comment robot
Since I personally have been doing self-media and publishing my works on various platforms, I am quite aware of this. Those erroneous comments are basically contributed by robots.
As for when it started, more than half of Internet content is produced by robots.Some reports suggest starting in 2022.
2022 is the year of the birth of generative AI. This year OpenAI launched the first large generative AI model-ChatGPT, which can directly imitate human creation content, and it is open to all ordinary users, and even some services are now free. .
The so-called robot-created content in the past is actually just a kind of “pseudo-original”. Some tools are used to make some simple modifications to the original articles – such as replacing some words with synonyms, so as to deceive search engines or social platforms. Get better search rankings.
However, with the emergence of the large generative AI model, everything has changed. Artificial intelligence can directly generate corresponding content in a few seconds based on your simple instructions. Articles, pictures, and videos can all be generated directly through instructions.
So, no surprise, it would be strange if most of what we see isn't produced by robots.
But not only that,Many of the so-called big Vs you follow on social platforms may be just robots behind them.
Social platforms do not disclose how much of their content is generated by bots, nor which accounts are bot-generated.
However, when Musk acquired the Twitter platform (now the X platform, similar to the domestic Weibo), the proportion of accounts run by robots on Twitter was brought to the table for discussion.
Twitter’s original boss told Musk that about 5% of daily active users (mDAU) were bots, but Musk asked a professional organization, Cybra, to conduct a re-evaluation. The company gave Musk two reports, one One report showed that 13.7% of Twitter’s daily active users were bots, while another showed 11%.
What’s even more funny is that these little more than 10% of daily active users of bots create most of the Twitter content.
You may be wondering why people let bots go to great lengths to produce content, while social platforms don’t filter it.
In fact, the answer is very simple. The platform cannot control it at all. The robots will constantly adjust their content according to the algorithm to avoid censorship by search engines or social platforms.
AI translation of Musk’s blog posts
But because the content produced by robots does affect the experience of real users, Musk is now exploring having Twitter users pay membership fees to block robot content (presumably in his eyes, robot content is as annoying as ads).
As for the purpose of producing this content, as far as I know there are three main points:
The first point is to earn advertising revenue from these social platforms.
Many people who have not been exposed to self-media may not know that when you publish works on the platform, you will get a share of advertising. This part of the revenue is now the main source of income for our IP.
When the market is good, a video with millions of views can earn several thousand yuan in revenue, but even now when the market is bad, this revenue can still be maintained.
Since popular content can easily continue to become hits, you will find that for a slightly more popular event or topic, you will encounter completely identical content created by different users.
Figure: AI generated pictures, which is currently very popular abroad, comes in various forms
The second point is account trading.
Those who work in self-media should all know that since the earliest interest in Weibo, an industry chain has emerged on the Internet that specializes in buying and selling social media accounts. This industry is considered a gray area on the Internet, but it has never stopped since its birth.
And from the initial sale and purchase of Weibo accounts, today we can buy accounts on any social media platform at very low cost, and accounts with millions of fans can be purchased easily.
Many people may want to know where the fans of these accounts come from. In the past, I always thought they were just zombie fans (inactive accounts), but in fact, even zombie fans have costs and the quality is still poor.
Many of these accounts are actually attracted by their content, but their content is not typed out word by word like ours, but directly picked up by robots in batches, modified in batches, and then published 24 hours a day.
Because it is completely produced and published by robots, it is very efficient. One operator can operate dozens of accounts at the same time and publish a large amount of popular content every day to attract users' attention.
When the fan base reaches a certain level, the account will be sold directly to make profits.
The third point, and the most annoying point, is that the perpetrators guide public opinion by publishing some content.
The comment area is a good place to guide the direction of comments, but in fact, robot creation content is not guided only by comments, but directly produces and publishes corresponding content.
Since most of the time it is difficult for us to understand the entire incident, it is easy to be misled by these out-of-context content. Some people with ulterior motives will use a large number of robots to publish “out of context” content to guide public opinion to the side that benefits them.
Therefore, any Internet melon must be eaten rationally, otherwise you may accidentally stand on the wrong side.
at last
If the purpose of the perpetrator is only to earn advertising shares and attract fans and sell accounts, then they will not be responsible for the content they publish. They will only care about how much traffic the content can get.
However, if the content currently created by robots is not reviewed by humans, it will only create false content, and use it out of context to guide public opinion as we mentioned earlier.
It can be said that there is a large amount of false information on the Internet today, so it is no exaggeration to say that the Internet is dead.
Reference links:
https://theconversation.com/the-dead-internet-theory-makes-eerie-claims-about-an-ai-run-web-the-truth-is-more-sinister-229609