Introduction
The New York Times recently hired its first AI editorial director. In this role, Zack Seward, a journalist with experience in product, digital media and news editing, will lead a team for iinnovate with AI tools, design training programs on AI and supervise the evolution of the industry.
The New York Times' Human Approach to AI
Despite this initiative, the publisher's stance on maintaining human-written news coverage remains unchanged. Seward will share their firm conviction that Times journalism will always be processed, written and edited by their expert journalists. It will also help determine how these new tools can help their journalists in their work, thereby increasing their reach and diversifying their reporting.
However, it is clear that the publisher will only use AI on its own terms. In August 2023, the NYT revised its terms of service to block AI tools from mining its content, and then blocked OpenAI's web crawler. In December, they sued OpenAI and Microsoft for allegedly replicating their work on ChatGPT and CoPilot.
The New York Times opinion on AI
Despite proposals from different AI companies for licensing deals and AI tools for journalists, they remained cautious, emphasizing the importance of human storytelling. Recently they have relaxed the rules slightly. In August, senior editorial staff considered how they could use AI to streamline their work.
Now, Seward and his team will be responsible for this process.
The State of Digital Publishing with AI
Each publisher seems to take a different approach to using AI. At the conservative extreme, we have publishers like the NYT and Wired who have taken a strong stance against the use of AI to write and edit articles.
In the middle of the spectrum, we have publishers like The Associated Press, which have used AI to generate short, to-the-point stories, like company earnings reports.
Publishers Turning Completely to AI
Finally, at the extreme end of the spectrum, trusting completely in AI, we have publishers like Channel 1. Set up for a full launch this year, Channel 1 reports real news, but its presenters are not.
The AI-generated actor in the video claims everything you'll see on Channel 1 is based on trusted sources and fact-checking, and uses AI to deliver news the way you want it: personalized, localized and condensed.
He adds in the introductory video that it was created to share unbiased, accurate and trustworthy news that upholds the central pillars of integrity and accuracy in journalism.
The Confidence Gap in Real-Life News Anchors
These fine words may be reassuring, but trust in real news anchors is already low. A recent survey found that only 42% of UK residents trust them, a drop of 16% year-on-year. So it's going to take a lot of marketing to convince consumers to trust AI-generated presenters.
Failures While Experimenting With AI
The Arena Group, which owns Sports Illustrated, Men's Journal and more, is another publisher that adopted AI early. However, their luck was not there. In February 2023, Futurism published an article detailing considerable errors in Men's Journal's first AI-generated article. Although the article claimed to have been reviewed by human editors, closer inspection by a medical expert revealed that it did not live up to expectations.
Consequences of AI Errors
“This article contains numerous inaccuracies and falsehoods,” Bradley Anawalt, director of medicine at the University of Washington Medical Center, told Futurism. “It misses many nuances that are crucial to understanding normal male health.”
The magazine also reported that CNET had more than 70 poorly written AI-generated articles on financial topics like savings accounts and payment apps. Almost a year later, Sports Illustrated was caught creating AI-generated content but passing AI authors off as real people. They strongly denied it.
The Interest of AI Companies in Publishing
As publishers try to figure out the best way to leverage the technology, AI players like OpenAI and Google are trying to monetize their work. OpenAI offers publishers millions to license their content and use it to train their large language models.
Apple, which is trying to catch up, is now doing the same. The company is offering more money for licensing deals with publishers such as The New Yorker, The Daily Beat, People, and Better Homes and Gardens.
A New Google Tool for Journalists
Last July, Google offered publishers large and small a tool called Genesis, designed to help journalists speed up the writing process. However, Google's new tool for generating search experiences is a big threat to the same publishers it is trying to woo.
As the world's most popular search engine, its new AI-powered search engine is expected to cost publishers up to 40% of their traffic, according to publisher estimates.
AI in Journalism: Friend or Foe?
So, with all these factors in mind, it still remains unclear whether AI is journalism's friend or foe, and under what conditions.
