There’s been a lot of talk recently about plagiarism and generative AI, most particularly but not exclusively around OpenAI’s ChatGPT. This is at least partially because such systems are trained on very large volumes of text or image data, and some ask if these algorithms are merely remixing prior art. But of course, the counter argument goes, how is that any different from how humans learn to write or draw?

The Verge’s Mia Sato tells us about controversy over CNET and generative AI in “CNET pauses publishing AI-written stories after disclosure controversy.” Futurism’s Jon Christian also writes about it in CNET’s AI Journalist Appears to Have Committed Extensive Plagiarism.

CNET, to their credit, has been using a generative AI system to write articles. Also to their credit, they have paused it due to potential problems. Futurism revealed the bot-written articles had errors and plagiarism. The AI’s text is mostly copied from human-generated raw material, so it can’t express new concepts, Christian claims. His article discusses examples of plagiarism in CNET’s generative AI system. However,, we don’t know what data the CNET system was trained on, and there’s no indication that it was based upon OpenAI’s technology (see below).

CNET held a one-hour call after they were criticized for using AI tools on the website. As CNET put it, they “did it quietly,” with a small disclosure on the website.The proprietary AI was built by CNET owner Red Ventures. Editors were able to pull data from domains and domain-level sections and source material to generate or augment stories.

During the call, leadership answered staff questions about the proprietary AI tool and gave details about its use. Interestingly, they did not answer questions about the dataset used for training, nor concerns about plagiarism.

They will now include disclosures about AI on their stories. They have also created an AI working group spanning across multiple departments, though it is currently unclear what the council has done so far.

Here are some of examples of what Christian describes as plagiarism. Granted, these are rather short and generic passages, so it’s difficult to imagine that there are many different ways to express the same ideas.

CNET original source

How to avoid overdraft and NSF fees. Overdraft fees and NSF fees don’t have to be a common consequence. There are a few steps you can take to avoid them.

How to Avoid Overdraft and NSF Fees. Overdraft and NSF fees need not be the norm. There are several tools at your disposal to avoid them.

Forbes Advisor
What is overdraft protection? Overdraft protection is an optional feature offered by banks to prevent the rejection of a charge on a checking account with insufficient funds. What is Overdraft Protection? Overdraft protection is an optional service that prevents the rejection of charges to a bank account… that are in excess of the available funds in the account. Investopedia

Becoming an authorized user can help you avoid applying for a card on your own, which is a major benefit if you currently have bad credit or no credit history.

Becoming an authorized user also lets you avoid having to apply for a card on your own, which is a major benefit if you currently have bad credit or no credit history at all.

 Bankrate

Here’s a screenshot of the original disclaimer on the CNET website:

Image 1 png

CNET’s Editor-in-Chief Connie Guglielmo has written about this, too, in CNET Is Testing an AI Engine. Here’s What We’ve Learned, Mistakes and All. They appear to be responsive to criticism. She writes:

In November, one of our editorial teams, CNET Money, launched a test using an internally designed AI engine – not ChatGPT – to help editors create a set of basic explainers around financial services topics. We started small and published 77 short stories using the tool, about 1% of the total content published on our site during the same period. Editors generated the outlines for the stories first, then expanded, added to and edited the AI drafts before publishing. After one of the AI-assisted stories was cited, rightly, for factual errors, the CNET Money editorial team did a full audit.

Their conclusions included:

  • These systems make mistakes—just like humans (no surprise there).
  • Disclosures and by-lines must be as visible as possible.
  • Better plagiarism checking should be implemented.

Author