As artificial intelligence continues to reshape education all over the world, understanding the impact it creates on teaching and learning becomes important. In an exclusive interview with The Free Press Journal, AI researcher, entrepreneur, and member of the UN Advisory Council for AI, Nazneen Rajani, discussed how AI transforms education from democratising access to learning to transforming the way of teaching. Rajani discusses the opportunities and challenges ahead as AI becomes an integral part of the educational landscape.
FPJ: How do you define AI in the context of education?
Rajani: Traditional education, where teachers assign homework and students complete it in a closed-book setting, is undergoing significant change. Today, many students are using tools like AI for their assignments. The focus is shifting from memorisation to applying information to make decisions and solve problems. Problem-solving skills are becoming increasingly important, and education is evolving to reflect this shift. It’s moving from a closed-book approach to an open-book format, where the emphasis is not on arriving at the correct answer but on the thought process and problem-solving approach, using all available resources, including AI tools like ChatGPT.
Moreover, education is poised to thrive with AI integration. There has been a significant disparity between those with access to education and those without. Access to education has been a clear determinant of better job opportunities, longer lifespans, and improved health outcomes. Conversely, those without access have lacked these advantages. AI’s democratization has the potential to bridge this gap. With AI, everyone has access to information and education, reducing the disparities inherent in traditional education systems.
It’s going to be a good moment for the education industry. There won’t be hurdles or obstacles – you don’t need to go to school to learn; you have school at your fingertips. Education will be more structured with AI’s help, with tools that personalize based on your learning ability, how you learn, what works for you – whether you learn through demonstrations or other methods. AI would work with that to help upgrade your skills.
FPJ: What would be the core benefit that students will get from using AI?
Rajani: The core benefit is that they will develop problem-solving abilities. Currently, our education has been rote-based and memorising-based, which hasn’t necessarily helped long-term. Many people ask where they use calculus in day-to-day life because they just tried to memorise derivatives without understanding how the world works or physics concepts. With AI, the focus will be on problem-solving. If given a task to build an application or solve a problem, it won’t be about retrieving information from memory but about using available information to solve tasks. Problem-solving skills will be much more developed for current students, and their opportunities will be better because people won’t be competing based on memorisation but on how they use information to solve problems.
FPJ: How can educators benefit from the integration of AI in education?
Rajani: Educators can focus more on real-world scenarios and prepare students for problem solving and real-world tasks. Currently, educators think from a traditional perspective, focused on closed book settings with assignments, quizzes, or homework. With AI being accessible, they’ll need to design courses, syllabi, and tutorials differently. AI will help them personalise learning for students. For example, some students learn slowly, and some students might not understand addition the traditional way. AI would help bridge that gap within a single classroom between students who are lagging and those who are advanced.
FPJ: What challenges might students & educators face with the use of AI in education?
Rajani: The biggest challenge for educators is that they need to upskill themselves. Many might see AI as a threat to their job. Instead of viewing it as a threat, they should see it as a tool to augment and help with their job, making it easier to personalise education for students. Since education will be heavily transformed by AI, educators need the right skill set. For instance, if a student misuses AI or if AI gives harmful advice and the student takes it, the repercussions could affect the educator. Educators need to understand AI systems’ limitations, which can only happen if they know how these systems work and when to rely on them.
For students, the challenge is they might lose interest in education, thinking ChatGPT knows and can do everything. Instead of using it as a tool to become more interested in learning and curious, they might lose interest, thinking, “Why should I learn when I can just use ChatGPT for everything?” This over-reliance is harmful because AI hasn’t surpassed humans as a whole yet. However, students with limited knowledge might think ChatGPT is already good enough.
FPJ: How do you suggest students avoid overrelying on it?
Rajani: Understanding AI’s limitations is important, which connects to educator training. When students are given these tools for educational purposes, they should be taught what AI can and cannot do. For example, you can use it to gain knowledge about certain things, but it won’t solve problems for you or your future job. We need to motivate them to understand this is a great tool, but it’s not magic—it’s not a silver bullet or magic pill to solve all problems. With AI penetrating different aspects of life, new problems will arise, requiring new skill sets.
In terms of education, we need to think about transforming our workforce to work with AI. People will need to learn how to work with AI agents. There will be teams of both humans and AI, so managing AI agents will become a skill itself. As we educate the next generation, we should think about teaching prompt engineering or using the right prompts from the beginning. Instead of banning ChatGPT, which students will use anyway, we could teach them to develop good practices so they’re not misguided or overreliant and use it beneficially.
FPJ: What skills will students need to develop to effectively work with AI in the future?
Rajani: In the next five years, there will be a mix of humans and AI agents. Working with AI agents will be important—assigning work and overseeing their work. There will be AI agents planning workflows and executing them, orchestrating given problems or tasks, determining steps, and collaborating on different things. Humans will be part of this, so we need to educate and upskill them on managing AI agents, understanding AI limitations, safety tenets, and when AI becomes a national security concern. This will require its own curriculum for working with AI agents. We’re still four to five years away from that, but since education is long-term, we need to start planning now.
FPJ: What emerging AI trends should students consider when pursuing a career in the field?
Rajani: Prompt engineering is important; knowing how to prompt and use AI will be important. Understanding limitations and what AI can and cannot do is also vital, depending on the type of AI—healthcare, legal, finance, etc. Each domain-specific AI has its own expertise, capabilities, and limitations. As students enter domain-specific majors in college, they’ll need to understand how to work with AI in their field, its limitations, and necessary safety guardrails.
FPJ: How do you address concerns about AI systems perpetuating biases in global education assessments or admissions processes?
Rajani: This relates to whether AI is the right tool for a given problem. Should we replace human admissions committees with AI? We’ve seen this issue before: Amazon used an AI tool that became discriminatory because it was trained on past data where they mostly hired male employees. It learnt a false pattern that female candidates weren’t good enough simply because the training data was mainly male. These biases are built-in because that’s how world data is represented. Fewer women go into software engineering compared to men, so AI made that judgement.
People are realising these are limitations of AI—it’s only as good as its training data. Now there’s more effort to ensure training data is well-represented and unbiased, representing world demographics fairly. Before using AI for every problem, we should consider if it’s the right tool. AI isn’t right for warfare, and currently, it’s not right for admission decisions because it’s still very biased. We’re working on de-biasing it, but it’s not there yet.
FPJ: What one piece of advice would you give students regarding the use of AI?
Rajani: AI is a tool with its flaws and limitations; it is just a tool. A pen is a tool; you can write with it or hurt someone with it. Similarly, AI is a tool. My advice is to use AI to be more productive, use it as a good tool, and not think of it as a threat that will take your job. AI is a tool, not a threat.
