'%20x='0'%20y='0'%20height='100%25'%20width='100%25'%20%0A%20%20%20%20%20%20%20%20%20%20xlink%3Ahref='data:image/jpg;base64,/9j/4AAQSkZJRgABAQAAAQABAAD/2wCEAAMDAwMDAwQEBAQFBQUFBQcHBgYHBwsICQgJCAsRCwwLCwwLEQ8SDw4PEg8bFRMTFRsfGhkaHyYiIiYwLTA+PlQBAwMDAwMDBAQEBAUFBQUFBwcGBgcHCwgJCAkICxELDAsLDAsRDxIPDg8SDxsVExMVGx8aGRofJiIiJjAtMD4+VP/AABEIAAUACgMBIgACEQEDEQH/xABgAAEBAQAAAAAAAAAAAAAAAAAABwgQAAEEAwADAQAAAAAAAAAAAAECAwQRAAUSBgcTFAEBAQAAAAAAAAAAAAAAAAAABQYRAAICAwEBAAAAAAAAAAAAAAECAwQAERIFFP/aAAwDAQACEQMRAD8AzPqfZMnTeGL1yYCHGp+v5QfpyWOyASkgWbrJNsNm5InynfywUfR9xXIjINdKurIxjE7U0khQM2wFGT3mVK9f6HiTlnmbo7J3n//Z'%3E%3C/image%3E%3C/svg%3E)
ChatGPT has been taking the internet by storm ever since it was released less than two months ago. The AI chatbot has been helping people of all ages, but students have benefitted the most by using ChatGPT to complete their assignments.
Recently, an AI model has been performing well on competitive and degree exams, scoring decent to high marks in various fields such as medicine and finance. This article will cover all the exams that the AI chatbot has passed so far.
Information about the examinations that ChatGPT has excelled in so far
ChatGPT is built on the GPT-3 model and has undergone extensive testing and development. The model has the capability to learn from previous conversations and user interactions.
However, the AI still struggles in certain fields like mathematics, and it cannot provide solutions for every query. Despite this, the model has been improving itself exponentially and will definitely become more capable in the near future.
1) Wharton's MBA exam
ChatGPT took an Operations Management exam assigned by Christian Terwiesch, a professor at Wharton Business School. After grading the paper, Terwiesch revealed that the model scored between B and B- (B minus), which is a respectable score for a student in the course.
Terwiesch further stated that the answers were not only correct, but the explanations were excellent as well. However, he was surprised that the model made mistakes during relatively simple calculations at an elementary level.
The current model of the AI does not seem capable of solving advanced mathematics. He added that the AI has the capability to modify its response when provided with hints regarding the correct answer.
2) United States Medical Licensing Examination (USMLE)
The United States Medical Licensing Examination is regarded as one of the toughest medical examinations in the world. It is a three-step exam that requires over four years of learning to pass. The tests contain difficult questions that cover topics ranging from basic science to medical management.
ChatGPT managed to clear all three examinations with over 60% marks and an accuracy of 50% across most of the questions. The researchers were impressed to see that the bot managed to achieve this feat without any specialized training.
3) Multistate Bar Exam (MBE)
ChatGPT had a 50% accuracy score while attempting the multiple-choice questions of the Bar Exam. The MBE is the first of three examinations in the Bar Exam that makes a student eligible to practice law.
However, the AI was only tested on the Torts and Evidence portions, which are two out of seven subjects. The bot correctly answered 50% of the questions without any prior training or knowledge. The researchers stated that a large and developing model would perform even better in the future.
ChatGPT is continuously improving itself with every query that it receives. The AI attempted questions from the latest exams, which were conducted after January 2022 and were not indexed on Google. OpenAI further stated that ChatGPT had been designed and trained on data available up until 2021.
Currently, the chatbot does not have access to the internet and relies solely on its training and the context of the conversation. It is believed that the chatbot's performance will improve once it has access to the internet.
