After Futurism found major errors in an AI-written “fact-checked” CNET article, CNET updated almost all its AI pieces as currently being reviewed “for accuracy”
Futurism Jon Christian
Related Coverage
Discussion
-
@jon_christian
Jon Christian
on x
CNET added this gigantic correction to one of its AI-generated articles after we reached out with some questions about its accuracy 🫠 https://twitter.com/...
-
@kpottermn
Kyle Potter
on x
Handing the reins over to AI to write content that could cost people serious money, lead them into credit card debt or financial ruin is a choice. https://twitter.com/...
-
@jon_christian
Jon Christian
on x
@Kantrowitz I think a human writer would have gotten a “hey man are you okay? is there anything we can do to help you avoid all these errors?” approximately 70 drafts ago
-
@tedgioia
Ted Gioia
on x
CNET is now letting an AI write error-filled articles for its site—but is trying to keep it secret. Welcome to the future of journalism. https://futurism.com/...
-
@farhip
Paul Farhi
on x
Turns out robots aren't much better at reporting than humans are, as a tech site has learned. @CNET's AI disaster. Latest by me: https://www.washingtonpost.com/ ...
-
@stefthepef
Stef Schrader
on x
Just hire real people, for Pete's sake. Hire people! Actual humans who know the subject matter. https://futurism.com/...
-
@matthewkeyslive
Matthew Keys
on x
Red Ventures' CNET said its AI-generated articles go through an editorial review process before they are published. But some articles contained so many errors that they've now been put under a separate review. https://futurism.com/...
-
@parismarx
Paris Marx
on x
Good analysis of the CNET's articles “written” by an AI. @Jon_Christian found the financial advice was bad and often wrong, setting up people with low financial literacy to be even worse off. The whole plan is all just to chase higher rankings on Google. https://futurism.com/...
-
@jakeswearingen
Jake Swearingen
on x
It seems inevitable that generative text AI models will get blended into journalism, but this is an excellent example at how to do it badly: 1) Don't be transparent 2) Use text AI for tasks it's not great at, like math 3) Let major errors from the AI slip past editors https://twi…
-
@techwontsaveus
@techwontsaveus
on x
“We couldn't help but notice that one of the very same AI-generated articles that Guglielmo highlighted in her post makes a series of boneheaded errors that drag the concept of replacing human writers with AI down to earth.” — @Jon_Christian https://futurism.com/...