A project launched by OpenAI fools readers into thinking a blog was written by a person
In the near future, artificial intelligence (AI) machines and robots will become more complex and will take over several tasks currently perform by humans. Writing seems like a task that requires more of what humans have and that machines don’t – creativity. However, the latest generation of a language prediction model presented by an AI company has demonstrated once more the potential of these AI-based bots. Developer Manuel Araoz working for the AI company OpenAI had the latest version of this bot to write an article about itself, and no one noticed the text was not written by a human.
According to a blog post by Araoz, the company released its GPT-3, the third generation of its language prediction model that is capable of creating “random-ish sentences of approximately the same length and grammatical structure as those in a given body of text.” The post also provided practical information about how this technology could be used to impersonate recognized figures by simulating their writing styles; like what Araoz did when it created a fake interview with Albert Einstein. In the future, this technology could be potentially replacing journalists, political speeches writers an advertising copywriters, according to Araoz.
The ability this bot has to predict sentences was used to make posts on the bitcointalk.org forum in the past few days. Overall, the posts received “positive” feedback that concluded that “the system must have been intelligent.” Regarding these posts, the blog said, “There are lots of posts for GPT-3 to study and learn from. The forum also has many people I don’t like. I expect them to be disproportionately excited by the possibility of having a new poster that appears to be intelligent and relevant.” However, that was written by GPT-3 and not by Araoz. “This article was fully written by GPT-3. Were you able to recognize it? This blog post is another attempt at showing the enormous raw power of GPT-3,” wrote GPT-3.