● 08.09.23
Gemini version available ♊︎
●● ChatGPT’s Decline is Continuing and New Studies Show That It Sucks (Pure Hype)
Posted in Deception at 12:12 am by Dr. Roy Schestowitz
Video download link | md5sum b99977573610aea94800f2ed9fdf9cb2Wasting Time With LLMs Creative Commons Attribution-No Derivative Works 4.0
http://techrights.org/videos/llm-hype-vs-reality.webm
Summary: Microsoft’s months-long effort to distract from mass layoffs with something it calls “HEY HI” (AI) but isn't may be coming to an end (facts overcome perception management)
THE seemingly endless and paid-for hype in the media created a bubble; companies rebranded as “HEY HI” (AI) to attain imaginary valuations, based on products that did not exist or were falsely advertised. Now, months later, the hype is gone and the bubble is imploding (ChatGPT is losing users at a rapid rate).
ChatGPT was never impressive; it was just marketed as “impressive” by media that got paid to say it was impressive, causing a sort of “fear of missing out” (FOMO).
ChatGPT will probably never even get much of a mention next year. What about other glorified “brands” from Microsoft/’Open’AI? Well, try to name one. Which one other than ChatGPT? In terms of usage, it’s going down and down every month. Right now it’s in the red again:
Image: SimilarWeb OpenAI Way past its peak
As it turns out, based on research from Purdue, there’s not only a disinformation issue when it comes to text but also code. To quote:
>
>
The Purdue team analyzed ChatGPT’s answers to 517 Stack Overflow questions to assess the correctness, consistency, comprehensiveness, and conciseness of ChatGPT’s answers. The US academics also conducted linguistic and sentiment analysis of the answers, and questioned a dozen volunteer participants on the results generated by the model.
>
“Our analysis shows that 52 percent of ChatGPT answers are incorrect and 77 percent are verbose,” the team’s paper concluded. “Nonetheless, ChatGPT answers are still preferred 39.34 percent of the time due to their comprehensiveness and well-articulated language style.” Among the set of preferred ChatGPT answers, 77 percent were wrong.
>
Well, that simple observation, and combined with the fact that it is plagiarism as a service, should remind people not to bother with ChatGPT, both for practical and legal reasons. Remember the ongoing class action lawsuit. █
Share in other sites/networks: These icons link to social bookmarking sites where readers can share and discover new web pages.
Permalink > Image: Mail
Send this to a friend
----------
➮ Sharing is caring. Content is available under CC-BY-SA.