Site icon Wonderful Engineering

CNET Started Using An AI Bot To Write Articles – But It Has Been Making Some Very Dumb Mistakes

Robot concept or robot hand chatbot pressing computer keyboard enter

CNET hoped to make a breakthrough by using an unspecified “AI Engine” of some sort, that wrote and published articles. With many different sites looking to use the same AI technology for a faster and better experience, the critics pointed out that this seems to be an experiment of some sort to eliminate jobs for entry-level and young writers. They argued that the accuracy and credibility of the AI text generators are inferior.

CNET editor-in-chief Connie Guglielmo, ignoring the riots, acknowledged the AI-written articles in a post that celebrated CNET’s reputation for “being transparent.”

But now if you happen to visit any of CNET’s AI-written articles, you can see an editor’s note at the top that says: “We are currently reviewing this story for accuracy. If we find errors, we will update and issue corrections.” This is because these articles are making horrendous mistakes and tons of factual errors. Even a great difference in quality can be seen in CNET’s previous articles and now machine-written pieces which seem more like just some facts thrown together.

Futurism noted that a story on CNET erroneously claimed that a person would only pay $1,000 in interest if they took out a $25,000 car loan with a 4 percent interest rate, ignoring the fact that payments are typically made monthly, and that interest is charged based on the remaining balance (so it would not be a flat rate).  This was checked with authoritative human resources with deep backgrounds in business who pointed out all the mistakes.

There were also many errors besides this one even though she promised that every story published under the program had been “reviewed, fact-checked and edited by an editor with topical expertise before we hit publish.”

All of this matters because AI has a great deal of promise to change the way we write, including journalism and PR. However, as we discovered in our experimentation with ChatGPT, the bots are often confidently and flagrantly wrong.

It’s up to us as humans to act as a defense against cocky machines. Errors are not new and human editors will always miss a thing or two. However, with a traditionally written article, you have multiple chances for people to catch errors: the writer and the editor.

If you incorporate AI into your writing routine, ensure you are editing the work meticulously and skeptically. The machine isn’t smarter than you —just yet.

Exit mobile version