Can artificial intelligence (AI) outperform humans when it comes to writing stories or articles? Different tech websites wrote about AI-generated blogs that fooled thousands of users. AI research lab "OpenAI" developed the GPT-3, the massive language model, which produces the articles.
The blog written by GPT-3 only showed how bad humans are at judging AI https://t.co/zCj18IxuMm pic.twitter.com/Fy7xswCjGr — Scott Feldman Esq. (@sfeldman0) August 24, 2020
Also Read: UPDATE: Elon Musk Neuralink Telepathically Starts Tesla; "Future is Gonna be Weird" Indeed
The AI community was excited ever since GPT-3 was released in July. Many developers received early access to the language model, allowing them to test how far AI research has come.
However, just like other AIs, GPT-3 received a lot of hype, while others question the claim as many publishers wrote stories that misrepresent the technology's capabilities. GPT-3's articles revived worries about robots deceiving humans, technological unemployment, and fake news attacks.
Liam Porr, a computer science student at the University of California, Berkeley, used the GPT-3 to create a Substack's blog under the pseudonym Adolos. However, Porr was not included in the limited audience of developers who have access to the tech.
To have access, Porr asked a Ph.D. student who already had access to the AI to run his queries. The student inputted a headline and intro in GPT-3 and received a full article.
Porr copy-pasted the best of several outputs of the AI model into his blog with minimal editing. His first article is titled "Feeling unproductive? Maybe you should stop overthinking."
He stopped the experiment and posted his last article, which contains his confession about the AI. Porr also speculated that GPT-3 could change future writing.
The GPT-3-generated blogs received negative feedback
Although the student said in his last article that the blog received 26 thousand visitors in 2 weeks, several reputable tech publications' news used the term "fake blog" in their headlines.
How ironic that an AI, which will replace all writers soon, is recommending people write on a daily basis. A comment from the original article
"A college student used GPT-3 to write fake blog posts and ended up at the top of Hacker News" https://t.co/QZKF2tdfjH pic.twitter.com/YarEXB9k7Q — Artin (@Artin) August 20, 2020
However, the word fake could be wrongly used to describe the student's blog since he did not spread misinformation. Although he didn't use his real name, using a pseudonym is an acceptable practice among bloggers.
There's nothing wrong with using penname as long as it is not used for ulterior motives or causes harm. The article generated by the AI can't be considered as fake news since they come from researches also made by humans.
For more updates about new artificial intelligence, always keep your tabs open here at TechTimes.
This article is owned by TechTimes,
Written by: Giuliano de Leon.
