Currently, in staffrooms worldwide, there are few topics more divisive than the subject of AI in education; is it simply a shortcut for lazy students or a transformative tool for personalised learning?
My staffroom where I work in Spain is no different, so I conducted a very basic experiment to try to bring some clarity to the issue. This exploration revealed that both perspectives hold truth.
I believe recognising and embracing this dichotomy while also realising that educators inherently possess the skills to leverage these aspects is the key to encouraging wider acceptance among teachers.
I challenged my sixth formers to answer a complex essay question on trade blocks, a subject we had not yet covered. They were directed to the computer room, equipped with their textbooks, internet and access to ChatGPT, but strictly prohibited from discussing with peers or seeking my guidance.
“Educators inherently already possess the skills to leverage AI.”
In a twist, I distributed further materials unevenly: half the class received comprehensive “ChatGPT Prompt” booklets. (For those unfamiliar with the term, “prompt engineering” involves crafting questions for AI systems like ChatGPT to get the best responses in the shortest time by understanding how the AI processes data.)
The other half were given only the minimal directive to “Chat with ChatGPT.”
The students were allocated one hour for the task. Upon completion, they printed their essays anonymously. I collected these unidentified submissions, keen to test if I could discern their authors based on their writing styles alone.
The “prompt” group completed their essays efficiently, taking as little as 20 minutes and used the remaining time to complete other homework. Their essays were uniform and impersonal, making it impossible to attribute even a single essay to their respective authors, and none could orally defend their conclusions. But, even with the occasional “hallucination” (or error in simple terms), they were all sound essays and worthy of good grades.
“The industry of AI consultants has exploded onto the scene and is highly profitable.”
Conversely, the “resource-less” group, though initially overwhelmed, delved deeper into ChatGPT. Their essays were notably personal and creative, each reflecting the unique style of its author, allowing me to identify almost all creators. The quality varied: students less adept in critical thinking produced original but below-average essays, while those skilled in it created their most remarkable work yet. Some of the answers were truly exceptional, and crucially, all could defend their conclusions.
What does this mean?
A solid grasp of “prompt engineering” undoubtedly enhances efficiency but not necessarily understanding. Engaging with ChatGPT through two-way dialogue—though less rapid— deepens understanding and leads to more effective, high-quality outcomes- provided the user has strong critical thinking skills. The efficiency-effectiveness distinction is a complex and well-debated topic in business and now merits similar attention in education.
The importance of teaching AI for efficiency
It’s a common frustration among teachers: students are submitting work using AI without deep engagement or understanding, yet the results are surprisingly good. While scepticism is natural, we need to adopt a slightly more nuanced response. In a rapidly changing workforce, students who don’t learn to leverage AI for efficiency will undoubtedly be disadvantaged. Whether we like it or not, it appears that AI is here to stay, and it’s our job to prepare them for it.
The importance of teaching AI for effectiveness
I feel this is a lesser-known topic, yet, in my humble opinion, the one that is most important for teachers to embrace. AI can become a personalised teaching assistant, enhancing learning experiences and developing students’ analytical abilities. AI’s potential in addressing Bloom’s 2 Sigma Problem is significant, bringing us closer than ever to personalised and more effective education for all. It really does have the ability to improve human intelligence.
Who will teach it?
Addressing who will teach AI in schools requires a shift in the current narrative. The industry of AI consultants has exploded onto the scene and is highly profitable. However, the cynic in me notes that these “experts’”now have a vested interest in maintaining AI’s inaccessibility to preserve their lucrative market.
For example, talk of “prompting” instead of “chatting” and “hallucinations” instead of errors, while technically sound, often creates undue obstacles and prevents a broader acceptance among less tech-savvy educators. Yet AI’s real value lies in its simplicity and user-friendliness. Indeed, this is why OpenAI has named it ChatGPT rather than PromptGPT.
“AI’s real value lies in its simplicity and user-friendliness.”
Spreading this message is crucial for the wider acceptance of AI among educators. You may be a technophobic humanities teacher who is worried about finding the time to complete the latest expensive online course about AI technology mandated by your school. But you might discover that, once you embrace it, your background in critical thinking and communication positions you better than most for using AI effectively – potentially even more so than the course’s instructor.