By Yogesh Bawane


There’s been a lot of talk lately about what teaching will look like in the age of artificial intelligence (AI). I know that many in my profession are reeling from the existential crisis of increasingly human-like AI. The ‘uncanny valley’ effect — or the unsettling feeling most of us experience as a response to androids or AR/VR simulations that mimic humans, but are not completely realistic yet — is a hotly discussed, but also widely joked about topic among robotics and AI practitioners. 


Given that educators are also responsible for constantly looking at everything from the lens of how it could help or harm a child. That ‘uncanny valley’ feeling now seems like a thing of yesterday, with AI almost perfectly mimicking human responses in many scenarios. 


So I get the 3am anxiety about human educators becoming redundant in the months and years to come. I’d be lying if I said I haven’t hyperventilated about it too – yes, even as a robotics and AI educator! 


My solution? Go back to the basics. It’s a lesson we teach all our students while nudging them to try a different approach to a problem they’ve been trained endlessly to think of in a particular way. 


And it’s this: Change is difficult and often uncomfortable. But it can also be exciting if we open our minds to its possibilities. 


Isn’t it funny how we sometimes forget to act on the very lessons we try so hard to teach our students?


AI is inevitable. And I don’t mean that in a doom-and-gloom Thanos-esque way at all! 


I simply mean that AI in education is the next exciting frontier, and if we focus on deploying it in an ethical and equitable manner, it can be the force that finally – at long last – closes the widening chasm between the ones with the most resources in our world and those with the least. 


Most of us instinctively and wholeheartedly believe that education is every child’s basic human and life right. And yet, here’s a technology that could help this right become a lived reality, and we’re focusing on all the wrong questions, instead. Or most of the wrong questions anyhow. 


Apart from the ideological and philosophical considerations, there’s also a very practical, everyday benefit to embracing the power of AI within the process of teaching to amplify the impact that human educators can have in improving learning outcomes. 


Actionable Feedback At Scale


Only a fool would deny that effective learning can take place without cutting-edge teaching. Is there a teacher on the planet who doesn’t aspire to avant-garde learning? But the reality of the matter is that with classrooms becoming bigger with more students per teacher and individual student time shrinking, most educators are over-extended. 


This is a necessary evil if the goal is to make learning more affordable and accessible to kids at the lowest possible price points. Learning is also a two-way process – teachers need to engage with students for them to actually learn what is being taught, but in a classroom of 20, can a human teacher be expected to start and engage in 20 simultaneous conversations on a topic? 


Neural networks, with their shocking ability to generate human-sounding text on almost any topic can help engage students in Socratic dialogue and prod them to look beyond their pre-set assumptions on a topic. And to me, that’s the whole point of teaching to begin with. 


This is a massive benefit even in 1:1 learning, because the ability to endlessly challenge and engage in an intellectual duel supported by AI helps teachers harness and expand the powers of their students’s creativity. I’ve witnessed this live in action in my own class. 


Sophisticated, Data-Driven Teacher Training


We often forget that while teaching is a calling, it is also a learned art – a muscle you have to constantly exercise for it to stay healthy and strong. So in many ways, good teachers are also forever students. 


Now imagine AI helping teachers get better at their craft by simulating students with different learning styles, providing real-time feedback that is rooted in the latest global innovations and best practices, thanks to deep learning from extraordinarily large data sets from around the world. Which teacher would not be a hungry student of AI? 


Focus On Creativity, Not Proficiency


Ask any educator and they’ll tell you that so much of a student and teacher’s time is spent on the pursuit of nailing the rules of a discipline. Like say grammar for instance. One school of thought is that spell check and tools like Smart Compose have robbed humans of the ability to remember spellings or form their own sentences. 


Another (which I belong to) is that now that students no longer need to memorize the millions of spelling rules within the language or use up their time or mental space in writing the same mundane filler lines that most functional writing entails, they are free to focus on higher-order tasks such as infusing originality into the ideas of what they are creating. 


For one, it’s a more efficient use of time. And, more importantly, the gatekeeping around writing suddenly vanishes now that it’s accessible to anyone with ideas, but lacks the ability to spell, for whatever reason. 


AI can help this transition in what we value – proficiency in rules, or creativity in what we imagine can be created. Every time I think of this question deeply, I’m reminded of the Danny Dunn series of 15 books, published in the 1950s and 60s, which deal with a young boy’s misadventures with science and technology. Danny Dunn, the young boy, is forever trying to invent things to either make his life simpler, just within the spirit of exploration or to get out of doing homework. And, satirically, his master strokes often come to him when he’s trying to get out of doing something boring and repetitive like writing punishment (which he gets a lot of) lines.


I could go on and on, but I will end by saying that it’s not that AI in education does not require deep reflection and examination. It’s just that we need to focus on better questions. Like how do we ensure that AI output authentically reflects the world’s diversity, instead of becoming another exclusionary tool by learning from and repeating only the narratives of those with unlimited access? Or how do we optimise models to engage in furthering the dialogue instead of delivering answers as soon as possible – I believe many world-class educational institutions have made significant strides on this front by training their models extensively. 


(The author is a Senior Educator of Robotics at BrightCHAMPS)


Disclaimer: The opinions, beliefs, and views expressed by the various authors and forum participants on this website are personal and do not reflect the opinions, beliefs, and views of ABP Network Pvt. Ltd.