AI is killing the old web?

Original source: CSDN

Image source: Generated by Unbounded AI‌

With the emergence of AIGC tools such as OpenAI's ChatGPT and Google's Bard, AI-driven content, information, and robots have been rapidly integrated into Web sites in the past year.

However, when these tools bring great convenience and improve efficiency, fake users, spam advertisements, and wrong messages also follow.

Just a few days ago, a social application start-up company IRL (IN REAL LIFE) was once valued at as high as 1.7 billion US dollars, and raised 170 million US dollars in the C round of financing led by Softbank. As a result, 95% of the company's claimed 20 million monthly active users are "bots," according to an internal investigation by the board.

NewsGuard, a company that provides trust ratings for online news outlets, has released a new analysis showing that, thanks to the ad budgets of major global brands such as tech giants and banks, they are flooding AI-generated low-quality sites with ad dollars. Websites with AI-generated low-quality content have received substantial financial support and continue to survive.

In this regard, foreign media The Verge reporter James Vincent concluded, "Nowadays, AI is killing the old Web, and the new Web is born with difficulty." And in his view, this is not necessarily a bad thing.

In just a few minutes, generative AI can create tons of text and images

Broadly speaking, years ago, The Verge pointed out, the Web used to be a place for individuals to create and innovate, and many people developed websites, forums and mailing lists on the Internet and made a little money from it.

Later, a lot of companies started and felt they could do better, so they created flexible and feature-rich platforms and opened the doors for more people to join.

It's kind of like, they put boxes in front of us, we fill those boxes with words and images, and people come and see what's in those boxes. These companies chase scale because once enough people get together there is usually a way to make money off of it.

But the emergence of artificial intelligence has changed these assumptions.

Today, when ChatGPT, Bard, and New Bing are popular, they can quickly generate a large amount of text and images, and can also make music and videos.

Realistically, their output may exceed the news, information and entertainment platforms we rely on.

Advertising revenue feeds a large number of AI-generated news and information websites

According to data released by NewsGuard, which has been tracking AI-generated news and information websites (UAINs) since May this year, 25 new UAINs can be found every week, and there may be hundreds of AI-generated content websites today.

So what is the point of these sites filled with a lot of AI content?

NewsGuard pointed out that one of the important points is that it has become an advertising place for many brand companies.

Analysts found through analysis that the advertisements placed on these UAIN websites are generated programmatically, which means that many well-known companies do not choose to place their own advertisements on UAINs, but are positioned by the advertising system. Most ads are served by Google Ads.

NewsGuard says their analysts have added 217 websites to its UAIN website tracker, many of which appear to be fully funded by programmatic advertising.

Since sites can make money through programmatic advertising, they have an incentive to publish content frequently. The company found that one of the UAIN sites published about 8,600 articles during the week of June 9 to June 15 this year, an average of about 1,200 articles per day.

In contrast, the news website "The New York Times" publishes about 150 articles a day, and the number of editors and reporters behind it is large.

In fact, in a recent study conducted in May and June of this year, NewsGuard analysts looked at sites in the US, Germany, France and Italy and found that 393 programmatic ads from 141 major brands appeared on 217 UAIN sites 55 of the .

Under such a trend, based on advertising revenue, a large number of new AI-generated news and information websites survive, flooding various places on the Internet.

Old Web site is being "baptized" by AI

In addition, a large amount of AI-generated content has also penetrated into many established communities and platforms, such as LinkedIn is using artificial intelligence to activate users; Snapchat and Instagram hope that when your friends are not talking, the robot will talk to you and so on.

At the same time, some Web sites are being passively impacted by ChatGPT, for example, Stack Overflow, a well-known programmer community, is one of them.

To this end, earlier this year, Stack Overflow also urgently issued a "ban", saying that it is prohibited to use content generated by ChatGPT to answer questions on Stack Overflow.

Because:

Contributions generated by GPT generally do not meet these criteria and thus do not contribute to a trustworthy environment. When a user copies and pastes information into an answer without verifying that the answer provided by GPT is correct, without ensuring that the sources used in the answer are properly cited (GPT does not provide this service), and without verifying that the answer provided by GPT is That trust is broken when the questions asked are answered clearly and concisely.

However, simply preventing the expansion of AI use is not the root of the problem. Stack Overflow plans to charge companies for scraping their data as they build their own AI tools for the use of AI. In addition, Stack Overflow itself has plans to provide AI-related services.

Another platform heavily influenced by AI is the question-and-answer community Reddit. "Reddit's data is valuable, but we don't need to give all of this value to some of the largest companies in the world for free," when Reddit CEO Steve Huffman said this, it meant that the company would do something.

Unexpectedly, when it updated the Reddit API charging policy, in order to get more income from it, and to make its data have a certain value, many moderators blocked the community forum to protest .

Finally, as the entrance to many websites, Google, the world's largest search engine, holds the lifeblood of modern Web sites. However, driven by the trend of Bing AI and ChatGPT as alternative search engines, Google is also taking action. According to media reports, Google is seeking to create a new search engine driven by AI technology. A major feature of the new engine is that it "can Guess what the user wants," the engine will learn and predict what the user wants to know based on what the user searches for, and will provide a pre-selected list of purchases, research and other information. If Google really implements this plan, the impact of AI on traditional websites will be huge.

** Let AI take the helm of the Web site, what will happen? **

There is no doubt that AI is now ubiquitous. If all platforms are eaten by AI, what will happen?

In this regard, Avram Piltch, editor-in-chief of the technology website Tom's Hardware, pointed out that although artificial intelligence has the ability to restructure text, it is people who ultimately create the underlying data-whether it is a reporter picking up the phone to check facts, or a Reddit user experiencing battery problems. I'd be happy to tell you how I fixed it.

In contrast, the information produced by AI language models and chatbots is often not necessarily correct, and even more troublesome is that when it is wrong, it is often difficult to find the wrong way. Therefore, it takes time and expertise to verify further.

If machine-generated content replaces the authorship of human beings, on the basis that only humans make mistakes, the addition of AI itself will make mistakes, and even hinder the ability of humans to absorb professional knowledge to a certain extent, then correcting our collective mistakes chances will be reduced.

At the same time, The Verge reporter James Vincent pointed out that the impact of artificial intelligence on the Web is not so simple to generalize. Even in the few examples cited above, there are many different mechanisms at play.

In his view, the most successful websites tend to be the ones that use scale to their advantage, either by increasing social connections or product selection, or by categorizing the vast swarms of information that make up the internet itself, but that scale relies on a large number of humans. Creating potential value, humans clearly cannot beat AI when it comes to mass production.

But is that necessarily a bad thing?

Not really, James Vincent argues, "Some would say it's just the way the world works, and point out that the Web itself kills what came before, often for the better. Print encyclopedias, for example, are all but extinct. , but I prefer the breadth and accessibility of Wikipedia to the thickness and assurance of Encyclopaedia Britannica. As with all things AI-generated writing, there are ways to improve it too—from improved citations functionality to more human oversight.Also, even though the Web is rife with artificial intelligence garbage, it may prove beneficial, stimulating the development of better-funded platforms. For example, if Google always Search gives you spam results, and you might be more inclined to pay for a source you trust and access them directly."

In fact, the changes AI is currently causing are just the latest in a long struggle in the history of the Web. Essentially, it's an information war—who makes it, how it's accessed, and who gets paid. But just because combat is familiar doesn't mean it's unimportant, nor does it guarantee that subsequent systems will be better than our current system. The New Web is trying to be born, and the decisions we make now will define how it develops.

reference: