ChatGPT and the Future of Learning

A lot has been talked and written about ChatGPT in recent weeks. Among the concerns expressed over the possibilities of this artificial intelligence (AI) tool are fears that we all might be out of a job soon and that students might now be able to cheat more efficiently on exams and assignments.

So what is ChatGPT? The short answer; it is a type of AI model known as a language model. It's been trained on a massive amount of text data, giving it the ability to generate human-like text and understand natural language input. It can be used for a variety of tasks such as generating text, answering questions, and having a conversation. This type of technology is used in chatbots, virtual assistants and other applications that require human-like language understanding and generation.

With regard to education, ChatGPT caused some stir when it recently passed both the bar exam and the US Medical Licensing Exam.[i] At Wharton Business School, professor Christian Terwiesch had the bot answer final exam questions from his “Operations Management” course and it received a B to B-, outperforming some students. Given these results, some have raised the question whether they are actually due to the ‘intelligence’ of the bot or rather based on how we convey knowledge and what type of learning we value.[ii]

Einstein on Education

Yet, this is not all-too new a debate. Philipp Frank, biographer of Albert Einstein, tells the story of Einstein visiting Boston in 1921. At the time, Thomas Edison was using a questionnaire to screen job applicants for his laboratory. Einstein, apparently, was not impressed by the questions. In fact, according to Frank, he admitted that he did not know the answer to one particular question, namely “What is the speed of sound?”. His reasoning was that he did not have to trouble himself with memorizing this specific fact as it is readily available in books. And, Einstein added “The value of an education in a liberal arts college is not the learning of many facts but the training of the mind to think something that cannot be learned from textbooks”.[iii]

In line with this view, in 2013, Neil deGrasse Tyson published the following tweet “When Students cheat on exams it's because our School System values grades more than Students value learning”. This could easily be applied to today’s debate around concerns how students might use AI tools to more effectively cheat on exams or assignments.

This current debate, in my view, provides the opportunity to rethink education now. Instead of feeling challenged by a bot we should ask ourselves whether we put the right priorities on learning. Rote learning, the memorization of information based on repetition, might not be our best guide for the future of education – especially in the context of increased digitization. We should, instead, foster curiousity, creativity and critical thinking.

The Impact of AI on the Labour Market

In a recent book, MIT professors David Autor, David Mindell and Elisabeth Reynolds explain the changing skill requirements caused by ongoing digitization.[iv] With digital technologies taking over more and more tasks previously primarily performed by humans, skill requirements for humans change. This is not a new phenomenon. Rather, new inventions or innovations have always had this impact on the labour force. This does not need to be a concern as long as human skills are a complement to what machines are able to do. Rather than being replaced by a machine, then, humans end up working with machines. In the best-case scenario machines can take over tedious, dangerous or repetitive tasks leaving the fun stuff to humans. The good news: as Autor et al. show, around 60% of jobs done in 2018 had not yet been ‘invented’ in 1940. Put differently, while technological change might destroy some traditional jobs, it also creates new ones.

Rethinking Education

What this requires, though, is a workforce able to adjust to the changes caused by technological change. And this is where education is crucial – not only for the current cohort of students. A critical feature in today’s labour market is going to be lifelong learning to enable workers to navigate the changing nature of work. This is where a changing paradigm to education comes in. First, with AI increasingly allowing us to get available facts in a fast and uncomplicated manner, Einstein’s dictum that education should focus on training the mind to think something that cannot be learned from textbooks, in other words the development of a creative mind, becomes even more prevalent than it was a century ago.

Second, learning should be fun. With increasing life expectancy we’ll all be more likely to experience disruptions throughout our work life. The ability to adapt to new situations and lifelong learning are imperative in such an environment. As such, education should teach us to stay curious and we should enjoy learning new stuff. Interestingly, bots like ChatGPT might have a potential application here - especially in the field of e-learning. With the ability to generate high-quality, engaging content, ChatGPT could be used to create online courses and educational materials that are tailored to the needs and preferences of individual learners. This could make e-learning more accessible and effective, especially for those who have difficulty with traditional learning methods.

Yet, as language models are trained on existing data sets algorithmic bias can occur. In testing AI tools that seemingly provide objectivity with regard to, for example, recruiting, online ads or facial recognition, researchers have repeatedly found racial, gender and socio-economic biases.[v] Hence, as a final point, teaching critical thinking is important.

Endnotes

[i] https://www.dailywire.com/news/chatgpt-passes-medical-license-exam-bar-exam-after-top-performance-on-wharton-mba-final

[ii] Max Muth, “Eine Zwei in Mathe für die KI”, Süddeutsche Zeitung, 24. January 2023.

[iii] Philipp Frank Einstein: His Life and Times (1947), page 214.

[iv] Autor, D., Mindell, D. A., & Reynolds, E. B. (2019). The Work of the Future: Building Better Jobs in an Age of Intelligent Machines. Foreword by Robert M. Solow. MIT Press.

[v] Lee, N. T., Resnick, P., & Barton, G. (2019). Algorithmic bias detection and mitigation: Best practices and policies to reduce consumer harms. Brookings Institute: Washington, DC, USA.