Posts Tagged: AIwritten

Microsoft retracts AI-written article advising tourists to visit a food bank on an empty stomach

Microsoft reportedly published — and retracted — an AI-generated article that recommended people visit a Canadian food bank as a tourist attraction. The article “Headed to Ottawa? Here’s what you shouldn’t miss!” included recommendations for catching a baseball game, honoring fallen soldiers at a war museum and… swinging by the Ottawa Food Bank. Paris Marx first called out the story on X (formerly Twitter). “People who come to us have jobs and families to support, as well as expenses to pay,” the AI-written section about the food bank section read. “Life is already difficult enough. Consider going into it on an empty stomach.”

Before its retraction, the article appeared on Microsoft Start, the company’s AI-aggregated news service that replaced Microsoft News in 2021. After The Vergereported on the article and its highly inappropriate recommendation about “going into it on an empty stomach,” Microsoft senior director Jeff Jones told the publication, “This article has been removed and we are investigating how it made it through our review process.”

The original URL now displays the message, “This page no longer exists. A new search page will load automatically.” The Verge uploaded screenshots of the initial story to Imgur.

The article’s author was listed merely as “Microsoft Travel,” suggesting real people may not have had any involvement in its creation. Microsoft Start’s “About Us” webpage claims it uses “human oversight” for the algorithms that “comb through hundreds of thousands of pieces of content sent by our partners” to help the company “understand dimensions like freshness, category, topic type, opinion content and potential popularity and publish according to user preferences.” The Windows maker reportedly laid off around 50 reporters from the division in 2020 while shifting to AI-generated news.

Microsoft is hardly the first company to get overzealous in its use of AI-created content. Early this year, CNET published numerous error-ridden financial explainer articles composed by artificial intelligence. More recently, Gizmodo’s parent company G/O Media posted an AI-composed (also mistake-filled) Star Wars article on the site, which deputy editor James Whitbrook called “embarrassing, unpublishable, disrespectful.” As the Associated Pressproceeds with measured caution on AI-assisted news coverage, other media outlets — including Microsoft’s news publishing wing — appear considerably more comfortable cashing in on fully AI-written articles, clearing the inevitable wreckage after the fact.

This article originally appeared on Engadget at https://www.engadget.com/microsoft-retracts-ai-written-article-advising-tourists-to-visit-a-food-bank-on-an-empty-stomach-182701884.html?src=rss
Engadget is a web magazine with obsessive daily coverage of everything new in gadgets and consumer electronics

CNET pauses publication of AI-written stories amid controversy

CNET is halting its use of AI-written articles for the time being. The Vergeclaims the technology publication's leadership has paused experiments with AI stories "for now" during a question-and-answer call with staff. While there's no word on the exact reasoning behind the freeze, which also affects Bankrate and CreditCards.com, editor-in-chief Connie Guglielmo reportedly said future AI-related stories would include a disclosure that the publication uses automated technologies.

Executive content VP Lindsey Turrentine also promised more transparency regarding the AI, according to The Verge. Some employees would get a preview of the tech, she said. More details of the system will reportedly be available next week. CNET owner Red Ventures has also formed an AI working group. Staff were generally unaware of either the AI's inner workings or when it was being used.

Questions about CNET's AI practices began last week, when The Byte noticed that dozens of financial explainer articles appeared to have been written using "automation technology." While there was a disclosure, it was effectively hidden when you had to click the byline to see it. CNET claimed in the blurb that humans "thoroughly" edited and fact-checked the work, but that wasn't true — the outlet started reviewing the pieces after Futurism discovered serious errors in a story.

CNET has used machine-made articles in years past. AI has advanced since then, though, and the discovery comes as text generation tools like ChatGPT draw flak and even bans over fears of plagiarism and reduced work for human writers. As with automation elsewhere in the workforce, some people don't trust that companies will use AI in an ethical way.

Engadget is a web magazine with obsessive daily coverage of everything new in gadgets and consumer electronics