Understanding GenAI

Artificial Intelligence (AI) refers to computer systems or algorithms that simulate human intelligence or problem-solving. It is not new technology – in fact, it has decades of use even in the field of education. In the 1990s, edtech companies used adaptive learning technologies to adjust the pace and style of digital content based on student needs, making learning faster or slower for students. In the 2000s, schools and colleges began using machine learning on their “big data” to predict student performance and intervention needs. Still, the release of ChatGPT in November 2022 caused a new flurry of interest in AI technologies.

Text-based interaction with AI (“chatbots”) and AI-powered image & video creation tools are powered by a subset of AI called Generative AI which produces content that often appears as though it came from a “human hand.” The graphic from AI for Education (2023) clarifies how GenAI is positioned in the broader field of AI.

It is important to understand that AI tools, GenAI included, are not inherently knowledgeable; they learn from large amounts of data collected from all over the internet and world, which introduces an inherent bias to their output. Imagine two 18-year-olds, one who grew up in rural Arizona and one who grew up in bustling Tokyo. They are both human, but their differences in lived experiences cause their worldviews and perspectives to be drastically different. Just as humans are influenced by both nature and nurture, a GenAI tool will reflect the algorithms and the datasets that trained them, both of which are selected by companies whose interests are not always aligned with those of educators. Further, GenAI models that currently exist cannot precisely define how or why they produce an output. Therefore, when it comes to biased or inaccurate output that feels “human” it is important to understand that the tool gives the statistically best answer or product based on what we ask, shaped by the training data used in that model.

The ability of GenAI tools to mimic human patterns of communication and creation has potential to be valuable in the world of education. For example, the use of “tutor bots” that allow students to have individual coaching that feels like a human conversation is being explored by both technologists and educators. Other education groups are attempting to use AI as a lever to reconfigure instructional time during the school day. As these explorations unfold, we will learn more about the impacts and net effect.

With a high regard for innovation and to set the tone of possibility, we continue with an investigation of GenAI’s role in powerful teaching and learning.