How to understand that a text on the Internet was written by an AI and not a human - ForumDaily
The article has been automatically translated into English by Google Translate from Russian and has not been edited.
Переклад цього матеріалу українською мовою з російської було автоматично здійснено сервісом Google Translate, без подальшого редагування тексту.
Bu məqalə Google Translate servisi vasitəsi ilə avtomatik olaraq rus dilindən azərbaycan dilinə tərcümə olunmuşdur. Bundan sonra mətn redaktə edilməmişdir.

How to understand that a text on the Internet was written by AI and not a person

Generative artificial intelligence is everywhere these days, including on the internet. Advanced text bots like ChatGPT can now write endless texts on any topic. They can make all this written content quite humanized, but it is still possible to distinguish a robot from an author, reports Gizmodo.

Photo: IStock

So how can you be sure that the articles you read on the Internet were created and published by a real person? While there is no foolproof or guaranteed way to do this, there are plenty of clues you can look for to determine what is AI-generated and what is not.

Check author

At least for now, there are no high-profile, reputable online publications distributing AI content without labeling it as such. But there are plenty of lower-level sites that take full advantage of AI-generated unlabeled text. If you come across a lot of text without attribution, this is one of the warning signs to look out for.

On the subject: A very strong humanoid robot has been created in China: it can help people at home and at work.

In contrast, if the article has a real person's name attached to it, along with a biography and social media links, then you're more likely to be reading something that was put together by a person. You probably won't have time to check everything you read on the internet, but it's worth it when you really need to know the source.

Supposed articles about artificial intelligence recently published on the Sports Illustrated website were accompanied by profiles and biographies of the authors - as it turned out, the profiles and biographies were also created using generative artificial intelligence.

Reverse image search (using something like TinEye) can reveal images of people that are not actually real, which can be useful in determining the source of an article.

Additional clues can be gleaned from the website in terms of its history, the type of content it publishes, the presence or absence of an About Us page, etc.

Use online detectors

There is a lot of debate about whether AI text recognition works. OpenAI says no, and most reports on this matter say that these AI detectors cannot be trusted. But within certain limits, they can be useful in testing the use of AI on the Internet.

Experts ran a short series of tests on several online AI detectors, including copyleaks, GPZero и scribbr, and what they found matches what other people have found. These detectors can distinguish AI text from human text, but only to a certain level.

These detectors seem to be more successful at detecting human writing than AI writing. Essentially, they look for originality in the text, trying to figure out what the AI ​​will say next based on its training. The more data they have to work with, the better, but there are limits to how much data you can use for free.

Research shows that some detectors are better than others, and some are even right most of the time, but none provide consistently high levels of accuracy. These detectors are perhaps best thought of as another tool that can be used alongside other research methods, rather than as something to be relied upon entirely.

Check the signs

There are certain signs to look out for: Due to the way generative AI is trained, its results tend to be general, vague, and obvious at times.

Certain notes of originality, humor and humanity are often absent (as are personal examples). AI always wants to generate text with a low level of confusion—in other words, a high level of predictability. Essentially, they simply predict what word should come next, and this can manifest itself in a general sloppiness and blandness that is sometimes noticeable.

You may be interested in: top New York news, stories of our immigrants, and helpful tips about life in the Big Apple - read all this on ForumDaily New York.

You can also look for obvious mistakes, but of course people make mistakes in their writing too.

By putting all these signals, hints, and flags together, you may be able to make an educated guess about the authorship of the text.

Read also on ForumDaily:

Virtual Disney World: how to visit famous amusement parks without getting up off the couch

The seven cheapest cities to live in America

Overtime work in the USA: when the employer is obliged to pay it, and when it may not

Miscellanea Educational program AI author's text
Subscribe to ForumDaily on Google News

Do you want more important and interesting news about life in the USA and immigration to America? — support us donate! Also subscribe to our page Facebook. Select the “Priority in display” option and read us first. Also, don't forget to subscribe to our РєР ° РЅР ° Р »РІ Telegram  and Instagram- there is a lot of interesting things there. And join thousands of readers ForumDaily New York — there you will find a lot of interesting and positive information about life in the metropolis. 



 
1084 requests in 1,196 seconds.