Artificial intelligence (AI) tools have already helped students cheat by generating research papers for them, and a new tool from OpenAI is no exception. The “Deep Research” feature, available to ChatGPT Pro subscribers, can produce long-form reports on various topics, including academic papers.
This tool, built on top of OpenAI’s o3 model, resembles Google’s similar “Deep Research” feature that was introduced last year. It uses AI agents to perform multiple internet searches and generate reports within 30 minutes or more. The Deep Research feature is designed to do the Googling for you, eliminating the need for students to decide on their own sourcing.
In a recent demo, OpenAI’s chatbot used this tool to analyze iOS adoption rates, language learning interest, and smartphone penetration. It cited sources such as Wikipedia in some instances, highlighting the potential for students to rely on this AI-powered research aid.
However, experts caution that relying on AI tools like Deep Research can lead to inaccurate or outdated information. The reports generated may not be based on the most up-to-date surveys from major analysis firms, and the AI’s interpretation of data may be subjective.
The process of using Deep Research takes longer than a typical ChatGPT request, ranging from under 10 minutes to up to 30 minutes. While OpenAI claims it can score higher than other AI models in tests, critics worry that its tools are being used for cheating purposes.
OpenAI is working on creating a version of the Deep Research tool that will be available to lower-cost subscribers, but most college students may not have access to this feature due to its high price tag.
Source: https://gizmodo.com/openais-deep-research-gives-students-a-whole-new-way-to-cheat-on-papers-2000558278