Openai is launching a new company that aims to give students a better understanding of complex topics. Just like the human introduction introduced in April, the learning mode will see Chatgpt in a Socrates way of conversation. Instead of answering questions directly, chatbots try to guide users to implement their own solutions, starting with questions that allow the system to calibrate their answers to goals and understanding. Then use the “scaffolding” structure for the conversation, which means Chatgpt will slowly roll out the information so that the user won’t overwhelm more information, rather than preparing to digest.
Openai said it has developed learning models in collaboration with teachers, scientists and pedagogical experts. Instead of running on a brand new model, the tool is powered by a series of custom system instructions.
“We chose this approach because it allows us to quickly learn from real student feedback and improve experiences – even if it leads to certain behaviors and mistakes in the conversation,” Openai said. “Once we understand the most effective approach through iteration and student feedback, we plan to train this behavior directly into our main model.”
It is worth noting that OpenAI is not a learning model that is only suitable for Chatgpt Edu users. Instead, the company first wants to launch the feature to log in to users for free, Plus, Pro and teams. EDU subscribers will be visiting in the “next weeks”.
It will be fun to find out how many students you end up using the learning mode, as the toggle allows you to turn on and off features easily. As the latest The article vividly explains that AI cheating is a major problem in American universities. Openai said it plans to work to make the learning model more attractive and useful to students. The company is exploring how to provide deeper personalization through the tool, as well as ways to provide target setting and progress tracking conversations.
