AI teachers can solve fundamental challenges in education today

In his 1967 novel The Vendor of Sweets RK Narayan introduced his readers to the idea of a ‘story writing machine’ that resembles a typewriter with buttons to choose the genre, characters, emotions and a few other attributes that can be pressed to automatically print out a story. One has to assume that the idea was to show the naivety of the character who did that. But less than six decades later, we may almost be there, thanks to technologies like ChatGPT and other such Artificial Intelligence (AI) driven tools that can not only write stories but also do other things like self-driving cars and robots in factories building other complex things. The jury is still out on AI replacing us for performing all the small and big things we are doing today, like driving or working in factories or even performing complex surgeries. But in many instances, like surgeries, we are already halfway there. So how about teaching too? An AI teacher may sound a bit much, even for the open-minded, but we may be closer to such an automated classroom than we think.

Since the possibility is on the horizon, the questions that need to be answered are why we should do it and what we can expect to change if we create AI teachers. Today’s education landscape presents a wide range of models, and

there is a clear gradation in quality based on the price one is willing to pay. From schools to colleges and other higher education institutions, the quality of education imparted can vary quite widely based on, among other factors the student-teacher ratio. The lower this number, the better it is for the student’s learning process and the quality of the outcomes. This premium value is largely driven by forces of supply, demand and affordability, meaning one can pay their way through it.

An AI teacher can become a great leveller or, shall we say, a democratiser of learning, meaning the opportunity to access high-quality education can become universal. Students will also have access to a teacher at any time of the day or, for that matter, any place, including one’s home. How we teach can also change quite dramatically. For example, some of the theories we learn in physics can be explained better on the playing fields, thus making the ‘content’ a lot more rich and interactive.

However, there will also be some uncertainties when we have technological breakthroughs like these. For example, if the teachers are, we may be quite certain that, apart from the subjects they teach, they will inculcate good values and ensure that the children are on the right path. One can argue if we are ready to let a machine raise our children in terms of the time they spend in school. For example, many of the values that we hold dear can be traced back to our experiences and learnings in school from teachers and peers. So, the question is if we can trust a machine to take the place of a human teacher.

One way to go about it is to start at the top by experimenting with higher education for working professionals, where the potential risk of introducing AI teachers is much lower and manageable. For example, when we are building a nuclear power plant, we understand it is better to be overcautious in terms of the safety mechanisms we build around it. It is better to go slow and very cautiously than rush into it because we do not know the long-term consequences of this shift, say 20 or 30 years from now. Again, to take the example of nuclear energy, we learnt this the hard way before we understood the full potential and risks involved in making it more relevant and human-friendly today.

In a country like India, a shift to AI teachers could potentially make a big difference by eliminating some of the most persistent challenges, such as the quality of education, inequality in access and opportunities and the massive demand-supply gap that we face in vital areas like higher education. More than eight lakh students appear for the IIT JEE exams for just about 18,000 seats available, and of late, even those who have graduated from such premium institutions are unable to find the work they had expected to land. With the speed at which AI is getting seeded in so many aspects of life, it might even be naïve to think it will not impact education in a very fundamental way. Maybe more naïve than RK Narayan’s character, who came up with the story-writing machine six decades ago.

This article is authored by Abhimanyu Saxena, co-founder, Scaler.