ChatGPT-4 – the computer interface you can interact with – just got even better. The AI chatbot uses natural language processing to create humanlike conversational dialogue and can compose written content in various modes, including essays, social media posts, emails, and exam answers. The new-generation language model is more advanced in three key areas: creativity, visual input, and longer context. Plus it’s getting better in writing exams.
OpenAI’s GPT-4 has been dubbed the “most-advanced” AI technology and it has passed numerous exams, demonstrating its abilities across a range of subjects. In comparison to its predecessor, ChatGPT, which is powered by GPT 3.5, GPT-4 can generate eight times the text and can comprehend and discuss pictures.
Here’s a list of the exams GPT-4 has passed, with marks, according to BusinessInsider.
Uniform Bar Exam
GPT-4 scored in the 90th percentile with a score of 298 out of 400, whereas ChatGPT only scored in the 10th percentile of the bar exam.
GPT-4 earned a score of 710 out of 800 for the Reading and Writing section, which puts it in the 93rd percentile of test-takers. For the Math section, GPT-4 earned a score of 700 out of 800, ranking among the 89th percentile of test-takers.
GPT-4 scored in the 99th percentile on the verbal section of the GRE, and in the 80th percentile of the quantitative section. However, GPT-4 only scored in the 54th percentile of the writing test.
USA Biology Olympiad Semifinal Exam
GPT-4 scored in the 99th to 100th percentile on the 2020 Semifinal Exam.
GPT-4 passed a host of Advanced Placement examinations, receiving a score of 5 on AP Art History, AP Biology, AP Environmental Science, AP Macroeconomics, AP Microeconomics, AP Psychology, AP Statistics, AP US Government and AP US History. On AP Physics 2, AP Calculus BC, AP Chemistry, and AP World History, GPT-4 received a 4.
GPT-4 scored a 30 and 60, respectively, putting it between the 6th to 12th percentile of the AMC 10 and the 45th to 66th percentile of the AMC 12.
GPT-4 has also passed the Introductory Sommelier, Certified Sommelier, and Advanced Sommelier exams at respective rates of 92%, 86%, and 77%.
ChatGPT has even come close to passing the US Medical Licensing Examination (USMLE), according to a study published in PLOS Digital Health.
While GPT-4’s impressive exam results demonstrate its capabilities, it’s not without its limitations. ChatGPT, powered by GPT 3.5, has been used to generate essays and write exams, often passing, but making mistakes too. For example, ChatGPT made simple mistakes in calculations when tested with questions from a Wharton MBA exam.
ChatGPT has also been shown to produce some terrible answers that discriminate against gender, race, and minority groups, something which the company is trying to mitigate. One way to explain this issue is to point to the data as the problem, blaming humanity for the biases that are embedded on the internet and beyond.
GPT-4’s exam results highlight its advanced abilities and demonstrate its potential to transform various industries. However, it’s crucial to remember that AI is not infallible and that human oversight is still necessary to ensure accuracy and avoid errors.
Categorised in: Uncategorized
This post was written by Syngli