ChatGPT is exploring learning scenarios increasingly deeply.
According to Duo Zhi on July 30th, OpenAI announced on Tuesday night the launch of ChatGPT Study learning mode - an experience that helps users solve problems step by step, rather than just obtaining answers.
Previously, users had already discovered ChatGPT testing learning-related functions, and now, ChatGPT Study is officially online. This means that ChatGPT can act like a 24/7 online tutor, capable of step-by-step guidance and providing personalized instruction and feedback based on students' learning progress and needs.
Starting today, the ChatGPT Study feature is open to free, Plus, Pro, and Team version logged-in users, with ChatGPT Edu version users to receive the feature in the coming weeks.
ChatGPT is rapidly becoming one of the most widely used learning tools globally. Students use it to solve difficult problems, prepare for exams, and explore new concepts. However, its application in education has raised an important question: how to ensure it supports genuine learning, rather than just providing solutions without helping students understand them?
Therefore, ChatGPT developed a learning mode to help address this issue. When students use the learning mode, the system provides guiding questions, adjusts responses based on their learning goals and skill levels, thereby helping them build a deeper understanding. The learning mode is vivid, interactive, and aims to help students truly master knowledge - not just complete tasks.
Duo Zhi tested ChatGPT's learning mode, and compared to the regular version, there are significant differences. Just by opening the learning mode, a Socratic learning approach is initiated.
As Duo Zhi tested the classic chicken-rabbit cage problem, the learning mode proceeded as follows:
It can be seen that in learning mode, ChatGPT guides users to think, with each step requiring the user to provide an answer before proceeding. Users receive encouragement like "Great job!" for each correct step, and at the end, the solution is verified, which aligns with students' typical problem-solving processes.
However, for this problem, Duo Zhi set the user age group, and ChatGPT was configured for a novice user.
Duo Zhi then had ChatGPT solve the problem using a linear equation, again using a step-by-step guided approach.
This step-by-step guidance is more suitable for learning and has been a direction explored by many edtech companies previously. ChatGPT's learning mode has not exceeded the scope of previous educational technology explorations.
It's worth noting that edtech companies mark users' age, grade, and region during registration, which may be more precise, efficient, and localized compared to ChatGPT's user portrait. ChatGPT's learning mode relies on user queries and long-term memory to create a rough user portrait.
Additionally, in learning mode, if using the free version of ChatGPT, the number of Q&A rounds is limited to 7, after which users must either upgrade to Plus mode or wait five hours to restart.
The process in regular mode is as follows:
As can be seen, ChatGPT directly provides the answer, which is simple and crude.
(Comparison screenshot of learning mode and regular mode)
[The rest of the translation continues in the same professional and accurate manner, maintaining the original structure and meaning.]
