AI Literacy: Why We Must Start Educating Children (and Adults!)

It’s 2023, and almost everyone has heard of artificial intelligence (AI) by now, and many are using it without even realizing it.

But few people outside the scientific and technology fields genuinely grasp what it is and how it should be handled.

This lack of AI literacy is starting to draw the attention of some leading thinkers in the technology and educational fields for several reasons.

First, users are less likely to embrace a new technology if they don’t understand it or cannot leverage it meaningfully. But an even more serious concern is that the ease at which people can engage with AI — far more accessible than the PC or smartphones — makes it highly probable that they can do real damage to themselves and those around them without a solid grounding in what it is truly capable of doing.

The challenge at this point, however, is to figure out what level of AI literacy is necessary for non-technical users and then guide educational efforts to empower significant populations with that knowledge.

The Practical Skills of AI

In a broad sense, AI literacy is how users have acquired the skills and competencies to engage AI applications effectively. This is not as easy as it sounds, according to educational technology writer Matt Crabtree. First, it requires the ability to view these highly complex technologies critically, to understand their strengths and weaknesses, and then apply these characteristics in contextually relevant ways in terms of their design and implementation.

AI in Early Education

Building AI literacy should begin at the elementary school level, according to many experts. Young minds are often the most adaptable to new things, and the proper introduction to AI will likely be crucial as the technology becomes more embedded in our world.

Education Week’s Alyson Klein notes that, after working behind the scenes for decades, AI is now front-and-center for today’s school-age population with the introduction of ChatGPT and other models. Without the proper guidance, young people might not be able to properly assess the power of these tools, and that could lead to any number of negative consequences.

Klein added that while it may be tempting to direct AI knowledge to only select students with penchants for science or computers, that would be a mistake given its rising ubiquity in modern life.

Primary and secondary education is rife with scientific instruction — everything from how a light bulb works to how plants turn light into energy — even though only a handful of students will grow up to be electricians or botanists. At a minimum, all students should receive hands-on training with AI in controlled settings, augmented by frank discussions on the moral and ethical aspects of its use.

AI Misunderstandings

One of the biggest challenges at the outset will be to undo many of the myths and misperceptions that have arisen around AI. Fueled by decades of science fiction books, movies, and television, much of the public has been conditioned to view AI as an all-knowing, infallible overlord capable of wiping out humanity on a whim.

A recent paper by researchers at the Georgia Institute of Technology noted that a critical element in this effort is to counter the tendency for children, and many adults, to believe that others (not just AI but animals and even people) think the way they do and can therefore be influenced by the same means. AI may be intelligent, but its cognitive processes are very different from our own, and it is only capable of mimicking key characteristics like emotion and intuition.

Other factors like gender, socio-economic background, and even race can influence the perception of AI and must be considered when developing instructional programs to build literacy.

Men, for instance, might be more inclined than women to tinker with AI to push it in new directions, and they might have inflated ideas of their success. People of African descent might be more sensitive to bias in AI results, especially when models have been trained with data generated primarily by European societies.

Playful Instruction

As a practical matter, however, there is no reason we can’t build AI literacy the same way we build literacy in math, language, social sciences, or any other branch of knowledge. A recent study by researchers at the University of Hong Kong noted that most learning paradigms begin with playful experiences like games and songs (“Now I know my ABCs. . .”) before gradually moving up toward more formal instruction. At each stage, the education industry as a whole should develop curricula that provide the most effective instructional environment and also establish the quantitative and qualitative criteria on which students and teachers are evaluated for AI literacy.

One of the most effective ways to do this is to elevate AI literacy to the same fundamental level as reading, writing, and arithmetic, with agreed-upon essential competencies for understanding and applying AI to the human experience. At the same time, the broader educational framework should be continuously revised and updated to ensure it remains relevant as both the technology and our understanding of it evolves.

The Bottom Line

Literacy is not about mastering a particular subject. One can be literate without understanding all the subtle intricacies of a Shakespearean sonnet or the minutia of particle physics. Literacy is about developing a working knowledge that helps you function in your environment.

As our collective environment becomes increasingly digitized and automated, understanding the fundamentals of AI will become as necessary as understanding how to use the Web or a mobile phone. AI is likely to be easier to adopt than any previous technology, but that doesn’t mean it won’t bring trouble to those who don’t have a firm grasp of how it should and should not be used.