First of all, I want to ask you a question: Have you ever used an AI assistance tool?
Image Source | Interface News
In this "AI era", the threshold for writing a decent article is already very low. With the help of AI, we may not be able to compare to the writing of real writers, but it only takes a framework of ideas, and AI can expand it into a complete article.
Image Source | Wen Xin said
I don't know if you have noticed that many search results in search engines have been occupied by AI-generated content, and many of them are full of errors and omissions "spam".
In the face of the encirclement and suppression of these AI-generated "spam content", how can we stand out from the encirclement?
Why AI is starting to produce "spam"
What is "spam"?
Imagine a scenario where your assignment or paper is facing a deadline and your work progress is really "miserable", so you have to ask the AI to generate a paragraph based on your description to "fill up the number", which is a type of "spam".
If you've ever used AI to generate content for you, you'll find that AI can only generate a satisfying manuscript if you polish your description of what you're doing.
In a study published this month, scientists working on AI at several prestigious scientific conferences were counted on the frequency of words used in peer review.
The word "meticulous" was more than 34 times more frequent than the previous year, "commendable" was about 10 times more frequent, and "complex" was 11 times more frequent. These phrases are a favorite of modern large language models such as ChatGPT.
Image Source | Zhihu
In other words, researchers at a large number of AI conferences hand over their peer review work to AI, or at least write with a lot of AI's help, and the closer the deadline gets, the more pronounced this phenomenon becomes.
In the past, we said that Deadline was the first productivity, but now it is probably the "first productivity" of AI.
In addition to being forced to use AI by the deadline, there are also some self-media practitioners who love "traffic is king" and are editing low-cost and high-efficiency fake content through AI, which all have the same characteristics:
1. The headline is amazing, attracts traffic, and the number of clicks and comments is high;
2. The content is formatted, each article is different, but it is very similar;
3. Guiding comments to vent their emotions, but ignoring the search for objective facts.
Unnutritious and fake content is flooding the Internet with the "help" of AI, and the time cost is increasing to find the information you really need from these "spam content".
Is all this "spam" the fault of AI?
AI can become more and more intelligent, relying on AI trainers to help AI become "smarter" through language model training, and on the other hand, relying on AI to obtain information from the Internet through large model data to "self-teach".
Trainers input a large amount of text, pictures and other content to help AI continuously learn and improve its content generation capabilities, which also makes those spam creators who make profits by producing "spam content" to obtain traffic, also eyeing AI to help them cultivate "content farms".
Content farms usually refer to the use of software to create a large amount of spam content to get users to click, so as to improve their ranking in search engines and become more searchable. At the heart of the content farm is "posting a variety of writing topics", based on a set of algorithms invented by the company, to evaluate two things: first, what is the current hot topic on the Internet, and second, what kind of keywords are more likely to bring high advertising revenue.
Image Source | Tencent Cloud Developer Community
In fact, "spam" polluted the Internet not in a day or two, but now AI is helping these people to produce this content more easily. You only need to come up with a sensational headline that attracts traffic, and then input it into the AI tool, using the "hallucination" effect of large language models (that is, it can generate content that is actually untrue around the information entered by the user), and the "spam" content can be quickly produced, and AI will become an "accomplice" in the production of "spam content".
How to make reasonable use of AI in the content production process?
What we should understand is that AI is not only capable of producing "spam", but can even write poetry with proper learning and training.
In 1989, the well-known science fiction writer Liu Cixin invented an "electronic poet" software that can produce new modern poems at a speed of 200 lines per second, and produce verses like the following:
"The asteroid is called!!/ Around the solid, there are only gelatinous rice paddies / No, I don't want to fly!!/ I miss // Trigonometry is watched!/Around Andromeda, there is only a living giant river / No, I don't want to devour myself!!/ I precipitate // The dragonfly is pinched!/ Around the Orient Express, there are only bows and arrows that are screaming / No, I don't want to smoke!!/ I talk // The confinement is warned!!/ Around the sword, there is only time to squeak / No, I don't want to sleepwalk!!/ I rot. ”
AI will help to produce "junk content", and it can also be a good partner in our writing and learning.
1. If you want AI to improve, you have to improve yourself first
Sometimes, we always feel that the AI is not as smart as it advertises, and it may be that there is a problem with the prompt word.
The "3W1H" model is a good way to structure prompts, that is, "who, what, why, how": define the role and goal you want the AI to play, clarify the requirements and details, explain the background and purpose of the requirements, and give the desired results and red lines.
Before using AI, train yourself to give prompt words, and maybe you can get twice the result with half the effort by cooperating with AI.
2. Make it your "inspiration partner"
A journey of a thousand miles begins with a single step, and the most difficult step in writing a good essay is the first step: finding a good topic.
If you feel that your inspiration is a little lacking, and you can't find the right person to discuss with you to broaden your horizons, then you might as well chat with AI, and let it help you capture the moment of inspiration through "100 questions in the industry", hot analysis, topic extension, etc., and you will have an "inspiration partner" who is always online.
3. Can't use titles? Ask the AI
How much of a good headline can add to an article? You may never know until you come across a good headline.
If you're a "hard worker" at heading, then you might as well let AI help you out and work out a few different headline style requirements for it, and choose one that you like the most and fit the topic of the article, so that the situation of "writing for 1 hour, half of the headline" will not happen again.
AI produces a lot of "spam content", but the source is the abuse of AI technology, and the correct use of AI can play a huge role in producing higher quality content. In other words, the title of this article is generated by AI, how do you think the AI performs? Can you make the AI generate a better title? Let us know in the background private message.
Source: Science China website - Beijing Science Center
Audit expert: Liu Xiangquan, Director of the Department of Robotics Engineering, Beijing Information Science and Technology University