Home >Technology peripherals >AI >US media is worried: The summaries generated by ChatGPT are good enough, but what if readers don't read the news?

US media is worried: The summaries generated by ChatGPT are good enough, but what if readers don't read the news?

王林
王林forward
2023-04-08 23:31:121660browse

US media is worried: The summaries generated by ChatGPT are good enough, but what if readers don't read the news?

According to reports, the US news industry is viewing AI chatbots as a new existential threat. They worry that people will think the article summaries provided by chatbots are good enough and stop visiting their sites, losing readers and advertisers. However, some media executives believe that despite the potential threats, there are also opportunities. They are trying to stay one step ahead of changes in the industry to adapt to the evolution of how readers get information.

The following is the translation

When you ask Microsoft's Bing chatbot whether former US President Donald Trump has been indicted, its answer will make media executives feeling scared.

The three-sentence summary given by the robot seems to be very useful. It not only provides links to news media such as CNN and the Washington Post, but also asks in a friendly manner at the end: "Is there anything else you need my help with?" ?"

But media executives view this generative AI technology as a new existential threat. They worry that people will decide that the article summaries provided by chatbots are good enough and stop visiting their sites, thereby taking away their readers and advertisers, repeating the situation in the early days of Internet innovation. They are looking for a variety of solutions, including blocking content, enacting regulations, or making readers pay for their work.

Some publishers are trying to prevent AI chatbots from crawling their articles. They can edit a file called robots.txt to try to prevent robots from accessing their site.

It’s unclear whether these measures will be effective. Francesco Marconi, a former reporter for The Wall Street Journal and The Associated Press and now co-founder of artificial intelligence company AppliedXL, said that if AI chatbots collect information in the same way as search engines, "publishers should be able to control which parts of their content are visible and potentially included, while marking other content as restricted.” However, he also said: "The actual operation of these artificial intelligence systems remains unclear."

Some media executives say they cannot protect their content because the new chatbots obtain information from search results , and publishers dare not exclude articles from search engines. On top of that, they know very little about how this rapidly evolving technology works.

“What happens inside the machine is not transparent,” said Dietrich von Klaeden, senior vice president of public affairs at German media giant Axel Springer SE, which owns Politico and online publications such as Insider.

AI chatbots have attracted the attention of the business world since the launch of ChatGPT by artificial intelligence research company OpenAI in November last year. In recent weeks, Microsoft has launched a chatbot using OpenAI technology, and Google has released a system called Bard.

Instead of typical search results showing a snippet of an article or a link to a website, the AI ​​service can generate longer responses.

Google said in a statement that it was using Bard to "learn and gather feedback from a variety of stakeholders, including news publishers," and that it would "continue to prioritize enabling us to deliver news Publishers send valuable traffic and support a healthy, open online ecosystem."

A Microsoft spokesperson also said that the company intends to cooperate with news media. "We have been working with publishers and have great respect for the content they create."

OpenAI did not respond to a request for comment.

Media industry executives are trying to assess the potential impact of AI chatbots on their business. Industry group Digital Content Next, whose members include The New York Times, News Corp. and Bloomberg, has held more than 10 meetings on the issue in the past three weeks. Some executives say AI chatbots may force them to rethink their strategies for engaging readers and rely more on other methods, such as sending newsletters.

Publishers also believe that despite potential threats, there are also opportunities. They believe AI can help cut costs and make newsrooms more efficient. Dietrich von Clayden said: "This can make journalists more efficient and allow them to focus on news creation, investigation and commentary."

Media executives are pushing for changes to copyright law, say Companies such as Microsoft and Google should be required to prove they have the right to use articles for artificial intelligence. One unresolved legal question is whether AI chatbots meet the "fair use" provisions of copyright law.

Dan Check, CEO of Slate magazine, said: "If you use other people's content to train and reduce the value of the content, then I think that is an unfair use."

Many publishers also want a share of the advertising revenue generated by AI chatbots using news to generate summaries. At an investor conference in March, News Corp. CEO Robert Thomson said his company had begun discussions about charging fees from artificial intelligence companies. "It's clear that they're using our exclusive content, and obviously we should be compensated," Thomson said. Newsrooms not only face financial threats from artificial intelligence; The challenge of chatbots spreading misinformation makes it more difficult for journalists to win public trust.

Publishers and search engines have a long and complicated history. For years, print media have complained about digital media companies that steal readers from them by creating headlines that are easy to find on Google searches and competing for high rankings in search results. Today, many newsrooms are training reporters on how to write search-friendly headlines, as this remains an important source of traffic.

Today, news industry executives are trying to stay a step ahead of changes in the industry to adapt to the evolution of how readers get information. “There is a new audience that consumes more content than any other group,” Marconi said. "And this audience is not made up of humans, but made up of machines."

The above is the detailed content of US media is worried: The summaries generated by ChatGPT are good enough, but what if readers don't read the news?. For more information, please follow other related articles on the PHP Chinese website!

Statement:
This article is reproduced at:51cto.com. If there is any infringement, please contact admin@php.cn delete