ChatGPT's Exam Competence

Launched by OpenAI, ChatGPT is a chatbot that produces its own material using generative artificial intelligence. The bot has been used to create essays and write exams, frequently succeeding but also occasionally making errors.

Wharton MBA Exam :

Christian Terwiesch, a Wharton professor, recently tested the technology using issues from his operations management final exam, which used to be a required course for all MBA candidates. He then released the results. Terwiesch came to the conclusion that the bot did an "amazing job" answering questions about fundamental operations that were based on case studies, which are in-depth analyses of an individual, group, or firm and a typical method used in business schools to instruct students.

However, there were other cases where ChatGPT made straightforward errors in calculations that Terwiesch believed only needed math at the sixth-grade level. Terwiesch added that the bot struggled to answer more difficult queries that called for a grasp of how various inputs and outputs interacted. Terwiesch predicted that the machine would ultimately score a B or B- on the test.

Microbiology Quiz :

Alex Berezow, a Big Think executive editor and science journalist, used a 10-question quiz on microbiology to evaluate ChatGPT. The questions, according to Berezow, who also has a Ph.D. in microbiology, would be suitable for a final exam for college-level students. ChatGPT "blew it away," according to Berezow.

Law Professor :

Based on a recently released paper authored by four University of Minnesota law school professors, ChatGPT recently passed exams in four law school courses at the university. In total, the bot answered over 95 multiple-choice questions and 12 essay questions that were blindly evaluated by the professors. The professors ultimately awarded ChatGPT a "low but passing grade in all four courses," roughly equal to a C+.

US medical licensing exam :

Researchers tested ChatGPT on the three-part United States Medical Licensing Exam, which candidates for medical careers take between medical school and residency, and they released their results in a paper in December 2022. ChatGPT "performed at or near the passing threshold for all three exams without any specialized training or reinforcement," according to the paper's abstract. Furthermore, ChatGPT's answers displayed a high degree of concordance and insight. Final findings indicate that ChatGPT's large language models may have "the potential" to help with clinical decision-making and medical education, according to the abstract.

Google Interview :

In addition to passing professional exams, ChatGPT recently aced the coding interview for an L3 job at Google. ChatGPT aced the interview, which was regarded as one of the most difficult to pass. The position pays about $183000 per year. This brings the debate over potential job losses due to ChatGPT and the issue of whether AI can genuinely replace human components back into the forefront.

Amazon Interview :

A few weeks earlier, ChatGPT also passed the coding test for the Amazon interview, but there were a few issues with ChatGPT's responses to the coding-related questions from the employer. They had "buggy" execution and were ineffective. An Amazon machine learning engineer says that ChatGPT was able to provide the correct responses and even enhance the codes. I'm both anxious and interested to see how this will affect how we handle coding interviews, he said.

Civil Service Exam :

The Bengaluru-based Analytics India Magazine put it through the Civil Service exams, which included queries about geography, the economy, history, ecology, general science, and current events, to assess its competence. The Magazine requested ChatGPT to answer all 100 questions from UPSC Prelims 2022 Question Paper 1 (Set A). "Only 54 of them were correctly answered by ChatGPT," it said. Even though ChatGPT's knowledge is limited to September 2021, questions about current events were not correctly answered. ChatGPT, on the other hand, gave incorrect answers for non-time-specific subjects such as economy and geography.ChatGPT uses word prediction to create writing that resembles that of a person. ChatGPT is unable to conduct online searches, unlike most chatbots. Instead, it produces text based on word relationships that are predicted by internal systems.

"ChatGPT is incredibly limited but good enough at some things to create a misleading impression of greatness," as per Sam Altman, chief executive at OpenAI.

Singapore 6th graders exam :

In addition to the UPSC exams, ChatGPT allegedly performed horribly on a test intended for Singaporean sixth graders. According to The Straits Times, it received an average grade of 16 out of 100 for the three mathematics exams it completed. It was awarded no points for any test questions that mentioned diagrams or graphs because it was unable to comprehend or respond to them. However, ChatGPT also made errors with straightforward text-based queries. It responded that the response was 65,503 when given the numbers 60,000, 5,000, 400, and 3, according to The Straits Times.

Conclusion

So As it is said that chatbots are going to replace human workers in every field and it is believed that now the invention of ChatGPT is going to replace human engineers. Therefore giving some evidence to the statement mentioned by giving information about exams or company interviews cleared by ChatGPT until now. Although it failed to pass the civil service exam and Singapore 6th graders exam due to its weak point in maths and current affairs as it only knows incidents after 2021.