DIGITAL LEARNING
essential that students are taught how to critically evaluate AI output ; examples include the potential for hallucinations , bias and discrimination , and reasoning errors ( Fitzpatrick , Fox and Weinstein , 2023 ), while large language models have , at least historically , struggled mathematically ( Kiela et al ., 2023 ), so students must be educated about the dangers of outsourcing both doing and thinking to AI . In essence , AI must be seen as a tool for students and staff in the same vein as the internet ; output should be regarded as a suggestion , not ‘ the answer ’. Teaching students critical thinking and evaluative skills will be more important than ever , not only to critique AI output , but also consider whether AI is the most appropriate tool for the problem in hand ; as in science , intellectual progress should always be problemfocused , rather than methodology-driven .
AI is still in its infancy and is only going to get better ; it will continue to evolve in uptake , methodology , and application . Students will also become increasingly savvy in their use and understanding of AI and arguably we should not discourage this ; solving problems using the most appropriate tools available is an important life skill , so the longer educators resist AI , the more ground teachers and schools will have to gain on students . We should not underestimate the value of working with technological advances , rather than against them ; in this respect , I prefer to think of AI as ‘ augmented ’, rather than ‘ artificial ’, intelligence . The AI age reignites the question of the purpose of education ; if it is to help prepare students for their future lives , AI is likely to become central to this . The skills in demand today are unlikely to be the same as those of tomorrow , and we must prepare students for this transformed world ; almost 100 million people will be working with AI in some capacity by 2025 ( World Economic Forum , 2020 ). Pedagogy will always require human intelligence around its art and science , and AI can help liberate more time for rewarding parts of the profession by delegating mundane jobs to automation ; in essence , enhancing staff wellbeing at no cost to student progress ( Churches , Hall and Sims , 2022 ). Although many are understandably cautious , I strongly agree that “ AI won ’ t replace humans , but humans with AI will replace humans without AI ” ( Lakhani , 2024 ), and my personal belief is that AI represents a golden opportunity to improve teacher wellbeing by reducing workload , positively transform the educational experience , and prepare students for the continued evolution of the Fourth Industrial Revolution ( Wolfram , 2020 ).
5