CHATGPT’s New Learning Mode: A Mixed Bag for Students

OpenAI has rolled out a new “learning mode” for its ChatGPT language model, which aims to guide students through their studies using the Socratic method. This mode, available to most logged-in users, including those with the free version, encourages students to engage with open-ended questions rather than simply seeking direct answers.

According to OpenAI, ChatGPT can function as a synthetic tutor that might significantly enhance academic performance. Leah Belsky, Vice President of Education at OpenAI, noted, “When ChatGPT is invited to teach, it can notably boost academic results, but it can also hinder learning if treated merely as a question-answer machine.”

Despite this promise, the efficacy of Study mode is still up for debate. Though it can provide direct answers with ease, just a click away in ChatGPT’s interface, this convenience may be problematic for younger users. They are still developing their cognitive skills and might struggle to resist the lure of quick solutions.

While some might argue that students have always found ways to bypass engaging with coursework—think Cliff Notes—the personalized and immediate nature of AI chatbots could exacerbate these tendencies. Educators have expressed worries about the secretive use of AI by students, particularly with the rise of apps that can solve homework problems via images.

Previous reports highlighted studies from MIT showing that frequent engagement with AI chatbots notably affects users’ brains, especially among the youth.

The findings were startling. Those who used ChatGPT for several months showed reduced brain engagement and performed poorly on verbal and behavioral tasks. Initially, this group interacted with the tool to ask structured questions, but by the end, they were more prone to simply copying and pasting generated content.

In contrast, participants using Google Search displayed moderate brain engagement, while those in a “brain-only” group exhibited the most robust neural connectivity. These results indicate that depending on AI models like ChatGPT could harm cognitive functions over time.

OpenAI’s CEO, Sam Altman, likens the introduction of AI in education to the advent of Google and computers, viewing it as a means to help students “think better.” The company is collaborating with learning experts at Stanford University to evaluate AI tools’ effects on educational outcomes, particularly in K-12 settings.