This study is coming as Microsoft is reported to be planning to invest $10 Billion in OpenAI. OpenAI is valued at $29 Billion.
According to a new study, fake abstracts and research paper generated by OpenAI’s artificial intelligence communication robot ChatGPT fooled scientists into believing that they were real papers.
ChatGPT-generated abstracts were identified by raters 68 percent of all times, and fake abstracts by 14 percent.
ChatGPT was used to generate fake research abstracts by Northwestern University researchers and Chicago University researchers. These abstracts were based upon 10 real studies published in medical journal. ChatGPT was also fed with fake research.
ChatGPT generated research papers that were 100% unique and bypassed all plagiarism detection programs. However, only 8% of them adhered to the format required for scientific journals.
ChatGPT, the chatbot with artificial intelligence, has taken the world by storm since its debut on November 30, 2018. It is today considered to be one of the most advanced robots, capable of not only writing essays but also generating computer code.
Researchers called upon medical journals and conference to adhere to strict standards and use AI identifiers for research evaluation.
The study comes at a time when Microsoft is believed to be investing $10 billion in OpenAI. OpenAI’s value is estimated at $29 billion.
Microsoft will get a 49% interest in the company as well as 75% of its profits, until it sells the investment amount.