Cloudflare CEO does not examine AI Chatbots source links

Companies that develop artificial intelligence always make the point to say that they include links to web sites in the answers created by the Chatbots for users. But the CEO of Cloudflare revealed Matthew Prince Intuition The search for research that continues to decrease. He said that publishers are facing an existential threat, because people do not click Chatbot and depend more and more on the summaries of artificial intelligence without drilling in a deeper way.

The prince said Intuition 10 years ago, Google sent one visitor to each two pages that had crawled. Six months ago, the percentage was one visitor for every six pages, and now it became one for every 18 years. Openai sent one visitor to a publisher for every 250 pages it crawled six months ago, while Antarbur sent one visitor to every 6000 pages. These days, Openai sends one visitor to a publisher for every 1500 pages, while Antarbur sends one visitor per 60,000 pages.

People have become more trustworthy in AI Chatbots in the past few months. The problem of publishers is that they do not earn ads if people do not click on the links that lead to their web sites, and for this reason Prince encourages them to take action to ensure their compensation somewhat. PRINCE said that Cloudflare is currently working on a tool to prevent robots that get rid of content for large language models even if the web page has already has “no crawling” instructions. If you were going to remember, I informed many ports in 2024 that artificial intelligence companies were Ignoring the protocol excluding robots on websitesOr Robots.txt, files and take their content anyway to train their techniques.

Cloudflare was looking for ways To prevent scraps Since last year. But it was only in March only when Cloudflare was officially introduced Amnesty International MazeWhich uses the content created from artificial intelligence “to slow, dissipate and waste the resources of artificial intelligence crawl and other robots that do not respect” not crawling directions. ” It works by linking an unauthorized creeper with a series of pages created from artificial intelligence that is convinced enough, but it does not already have the contents of the site that protects it. In this way, the creeper ends with wasting time and resources.

“I go to war every day with the Chinese government, the Russian government, the Iranians, the North Koreans, perhaps the Americans, the Israelis, and all those who are trying to penetrate the locations of our customers,” said Prince. “And you are telling me, can’t I stop some of the student who studys a lot with Corporation in Palo Alto?”