 |
| OpenAI's head of education, Leah Belsky, says the company is developing ways to introduce "productive struggle" into AI education tools. Reuters |
Leah Belsky, OpenAI’s vice president of education, has a clear message for students navigating today’s rapidly changing academic landscape: AI is not here to do the thinking for you it’s here to help you think better.
Speaking on OpenAI’s official podcast on Friday, Belsky emphasized that students need to embrace artificial intelligence as a tool that enhances learning, not a crutch that replaces it. “People who use AI in the workforce are significantly more productive,” she said. “So it only makes sense that students should start building those skills before they graduate.”
Belsky, who has led OpenAI’s educational initiatives since 2024, believes the current resistance to AI in schools is misplaced. Rather than banning or restricting tools like ChatGPT, she says educators should be teaching students how to use them in ways that foster critical thinking and creativity.
Rethinking AI in the Classroom
In many classrooms today, AI usage is either discouraged or outright prohibited, largely because it’s seen as a form of academic dishonesty. Concerns abound over students relying on chatbots to complete assignments without doing the actual work, raising fears of cognitive stagnation what some call “brain rot.”
But Belsky sees it differently.
“AI is ultimately just a tool,” she said. “It’s like a calculator. What matters in education is how that tool is used. If a student uses AI as an answer machine, they’re not going to learn. But if they use it to explore ideas, expand their thinking, or generate creative outputs, then AI becomes a powerful educational ally.”
This is why OpenAI is actively working on ways to encourage more meaningful use of its technology in education. One example: the newly released “Study Mode” in ChatGPT, designed to guide students with tailored questions that match their learning objectives and skill levels. According to the company, this feature is meant to help students build deeper understanding, not just retrieve answers.
Coding: The New Literacy
For Belsky, another key part of preparing students for the future is ensuring they learn to code and learn how to code with AI.
“Coding is becoming a core literacy,” she explained. “Thanks to things like vibe coding, where you prompt AI in natural language to generate code, we’re entering a world where every student should know how to use AI to create apps, images, and software.”
Vibe coding, which allows users to describe what they want and let AI write the code, has become a popular method for beginners and professionals alike. However, the technique still requires basic programming knowledge. AI-generated code can contain errors, and users need to be capable of debugging or validating the results or at least know someone who can.
The Value of Productive Struggle
One of the biggest concerns surrounding AI in education is that it might eliminate “productive struggle” the process of grappling with difficult concepts that is essential to learning. Without that effort, students may absorb less and rely more heavily on quick solutions.
OpenAI acknowledges this concern and is actively trying to address it.
“Part of what we’re building is friction intentional resistance at the right points in a student’s learning journey,” Belsky said. “It’s about designing interactions with AI that require the student to think, reflect, and build understanding step by step.”
Other tech companies are exploring similar solutions. Kira Learning, a startup chaired by Google Brain cofounder Andrew Ng, is focused on helping teachers particularly those without a background in computer science introduce AI and coding into their classrooms. Kira has rolled out AI-powered agents aimed at creating meaningful learning experiences by introducing challenges at key moments in a student’s process.
Andre Pasinetti, Kira’s CEO and cofounder, told Truth Sider that the goal is to help students experience just enough difficulty to support real comprehension. “The trick is to create friction at the right moments so students actually learn through the experience,” he explained.
Shifting the Education Paradigm
While these innovations are promising, they highlight a broader truth: the education system is playing catch-up. Much of the responsibility for integrating AI into student learning currently lies with technology companies, not schools.
And according to some experts, academia needs to reevaluate what success in education really means.
Tyler Cowen, a professor of economics at George Mason University, recently shared his thoughts on this shift in a conversation with tech analyst Azeem Azhar. “There’s a lot of hand-wringing about ‘how do we stop people from cheating,’” Cowen said. “But the real question is, ‘What should we actually be teaching and testing?’”
Cowen warned that traditional academic incentives namely, doing whatever it takes to get top grades are becoming increasingly obsolete. “The system is still built around optimizing for grades, not for learning or creativity,” he said.
As AI tools become more advanced and more accessible, the education sector faces a defining challenge: adapting fast enough to prepare students for a future where AI is everywhere. That doesn’t mean replacing traditional teaching methods, but enhancing them.
Leah Belsky believes the answer lies in thoughtful integration. “AI isn’t going away,” she said. “So the question becomes: How do we help students use it responsibly so it makes them better learners, not passive ones?”
If used wisely, tools like ChatGPT can become as essential to modern education as the calculator once was. But only if students and their teachers know how to use them.
