According to reports, ChatGPT, an AI chatbot created by OpenAI, failed the Union Public Service Commission’s (UPSC) notoriously difficult Indian Civil Services Examination, which is regarded as one of the toughest tests in the world.
When ChatGPT was introduced in November 2022, it quickly became well-known for its success in passing numerous US tests, including the US Medical Licensing Exam (USMLE) and other MBA exams, as well as the Google Code Interview for Level 3 Engineers.
Insights India Magazine, a Bengaluru-based publication, put ChatGPT to the test by giving it all 100 questions from Question Paper 1 (Set A) from the UPSC Prelims 2022. These questions covered subjects like geography, economics, history, ecology, general science, and current affairs.
Only54 of the questions could be satisfactorily answered by ChatGPT, and it gave incorrect responses for subjects like geography and economics that were not time-sensitive.
ChatGPT was anticipated to perform better on current events queries, despite its knowledge cap of September 2021. The chatbot can’t perform internet searches like most chatbots because its goal is to create writing that resembles that of a human by predicting incoming word sequences.
Sam Altman, the CEO of OpenAI, has said that ChatGPT’s powers are constrained and that it might inadvertently convey a false sense of greatness in some contexts. Also, according to sources, ChatGPT did poorly on a Singaporean exam for students in the sixth grade.