Shift from Search to Conversation
Students worldwide are increasingly turning to artificial intelligence tools such as ChatGPT to find answers to physics questions, according to a new study led by researchers at Bar-Ilan University.
The study, published in Physical Review Physics Education Research, analyzed search data from more than 20 countries between 2022 and 2025. The researchers found a consistent decline in Google searches for physics concepts and a similar decrease in visits to related Wikipedia articles, coinciding with the rapid adoption of generative AI tools.
Dr. Yossi Ben-Zion and Omer Michaeli of Bar-Ilan University's Department of Physics conducted the study in collaboration with Prof. Noah Finkelstein of the University of Colorado Boulder. The research was based on Google Trends and page views for physics articles on Wikipedia in seven different languages.
"The findings do not suggest that students are learning less," said Ben-Zion. "Rather, they indicate that many students are shifting from searching across multiple websites to receiving direct, conversational explanations from AI systems."
Traditional online research often requires students to search across several websites, compare sources, and piece together an explanation. AI tools offer a different experience by providing a direct, conversational response to a question, which could make finding information faster and more accessible, particularly for students who may struggle to identify reliable educational resources online.
Language and Educational Equality
One of the study's most striking findings, as reported by Mirage News, is the difference between English-speaking and non-English-speaking countries. In the United States, the United Kingdom, and Australia, Google searches for many physics topics remained relatively stable. In contrast, many non-English-speaking countries experienced substantial declines.
The researchers propose that AI may be acting as an educational equalizer by allowing students to receive clear, high-quality explanations in their native language rather than relying primarily on English-language educational resources. Open Access Government also highlights language as an important factor, noting that much of the educational material available online has traditionally been produced in English, meaning students who speak other languages can face an additional barrier when trying to understand complex subjects.
"Generative AI has the potential to reduce the 'language tax' faced by millions of students," the researchers explain. "By making complex scientific concepts more accessible in many languages, these tools may help democratize access to knowledge."
Topic-Dependent Decline
The study also found that the transition to AI depends on the nature of the subject matter. Google searches declined much more sharply for physics topics that can be explained primarily through text and logical reasoning than for topics that rely heavily on diagrams, graphs, and other visual representations. This pattern was consistent across the countries studied.
Implications for Education
The findings offer one of the first large-scale international looks at how generative AI is transforming students' information-seeking behavior, according to the researchers. The study's implications extend to teachers, universities, and educational publishers, who may need to adapt to changing habits in how students access and consume educational content.
While the researchers emphasize that the shift does not necessarily indicate a decline in learning, the study underscores a significant change in how students engage with scientific knowledge, driven by the rise of AI-powered conversational tools.