ChatGPT successfully solved 80 percent of these intelligence-based problems, which closely resemble challenges found in standardized tests like the SAT—an exam widely used for US college admissions.
ChatGPT successfully solved 80 percent of these intelligence-based problems, which closely resemble challenges found in standardized tests like the SAT—an exam widely used for US college admissions.
ChatGPT, an AI language model developed by OpenAI in 2020, exhibits problem-solving abilities comparable to, or even surpassing, those of college students in the United States, according to a study conducted by psychologists at the University of California, Los Angeles.
The research, published in the prestigious Nature Human Behaviour journal, highlighted the capabilities of AI in tackling complex reasoning problems, the Erudera reported.
The researchers at UCLA decided to assess the predictive capacity of GPT-3 by presenting it with a series of intricate arrays of shades and analyzing its ability to predict the next image.
ChatGPT successfully solved 80 percent of these intelligence-based problems, which closely resemble challenges found in standardized tests like the SAT—an exam widely used for US college admissions.
In comparison, approximately 40 UCLA undergraduate students were also tasked with the same problems.
The results showed that while the AI model achieved 80 percent accuracy, the students performed below par, solving less than 60 percent of the problems presented to them.
Lead Researcher Taylor Webb highlighted that while ChatGPT demonstrated remarkable problem-solving skills, it encountered difficulties when tasked with tasks involving social interactions, mathematical logic, and certain problem-solving scenarios, such as transferring sweets from one bowl to another.
Webb pointed out that the model, although advanced in its own right, is not at the level of fully generalized human intelligence but has undoubtedly made significant progress in specific areas.
The introduction of ChatGPT into the education sector raised concerns among universities worldwide regarding potential student plagiarism, according to a survey conducted by Best Colleges.
As a result, many academic institutions opted to restrict access to the AI tool to curb cheating incidents. Some institutions even issued stern warnings to their students, advising them against using AI-based solutions to complete their assignments.
However, experts in the education sector maintain that while AI tools might facilitate the completion of assignments, they do not diminish critical thinking and problem-solving skills which are essential attributes for academic success.
Jenna Lyle, a spokeswoman for the New York City Department of Education, asserted that ChatGPT, despite its prompt and easy responses to various questions, fails to instill the vital skills required for achieving academic excellence.
Harvard University recently announced that starting in the upcoming fall semester, computer science students will be able to use AI to identify errors in their codes.
Harvard’s CS50 course will be taught entirely by a chatbot, demonstrating the growing influence of AI in the field of education.
