Google Discover tests AI-generated headlines, and the results raise questions
Google continues to integrate artificial intelligence into its services, even where it seems inappropriate. As discovered by The Verge, the company is testing AI-generated headlines in the Discover feed, which may not match the originals. Unfortunately, some of them are inaccurate or even misleading.
What is known
One example is an Ars Technica article about the Valve Steam Machine. The article's headline was cautious: "Steam Machine looks like a console, but don't expect its price to be so." However, in Google Discover, this material appeared with the headline "Steam Machine price revealed," although there was no pricing information. Similar errors were noticed by Engadget journalists, who also recorded that descriptions are sometimes added to original headlines. In such cases, Google adds a note: "Created with AI, which may be incorrect."
Google spokeswoman Mallory DeLeon explained that this is a "small experiment for a portion of users," aimed at improving the perception of the topic before a person clicks on the link. However, against the backdrop of complex relations between Google and the media, the situation looks ambiguous. The company has been repeatedly criticized for using other people's content without proper compensation. And when publishers demanded payment, Google sometimes simply removed their materials from the search.
Meanwhile, the integration of AI into Google Search is only increasing. Users can already try AI Mode - a feature that the media organization News Media Alliance has called "theft." Now Google is testing a new interface where AI Mode and AI Overview are displayed on one screen.
Source: The Verge
