Technology is advancing fast, and AI is moving into classrooms, but there are concerns about which might be better for education.

AI is increasingly incorporated into schools because of technological advancements and the growing desire for tailored education. While concerns must be tackled, the potential advantages are quite enticing. This article examines the perspectives and evaluates the advantages and disadvantages of implementing AI in early childhood education.

As more schools use AI, students, teachers, and schools must figure out how to use these tools safely. Some teachers are against chatbots like ChatGPT because they can help students cheat and spread false information. We know that AI chatbots may contain false or inaccurate information, even leading to nonexistent websites. Also, experts and consumers have asked important questions about AI’s data protection, algorithmic bias, and unequal access.

They worry about cheating and students not talking to each other enough in the classroom. Teachers also worry that students may have less contact with real people as AI is used more. It’s not like the CB radio days where people all over the country talked to each other—now they will talk to a computer. And not all AI chatbots have current information.

By analyzing the data provided to them, these programs are trained to find patterns, structures, and subtleties that they can then use to guess the right answer to the user’s question. When the AI system uses old data as input, it will show these mistakes in the results. AI systems do not have cognitive skills and cannot tell themselves if the information they use applies to the query.

Since “large-language model” (LLM) apps like ChatGPT and generative artificial intelligence (AI) came out, many people think that this new technology will completely change the way we teach. For example, Bill Gates said in April 2023 that generative AI “will be as good a tutor as any person could be” in 18 months. This is wonderful but over-promising, as we’ve seen in many recent articles published in scientific literature.

The International Scientific Report on the Safety of Advanced AI is now out. General-purpose AI, both now and in the future, can be used like all strong technologies. AI can spread false information on a large scale and impact operations, fraud, and scams. General-purpose AI that does not work right can also be harmful, like making biased choices based on protected traits like race, gender, culture, age, or disability.

Personally, I’ve received data that assumed all users were men. Who did most of the programming, and what were their biases? I recall a young boy who once, when a woman said she’d like to be an engineer, responded, “Girls can’t be engineers. Only boys can.” Today, the young boy is a nurse. Case closed on that one. I’ve also been directed to non-existent websites or articles that don’t exist; AI made them up.

But people like Sal Khan, who founded The Khan Academy, believe in the technology’s ability to improve education. However, he recognizes where improvements in AI technology are needed.

Elementary education is the most important educational component in a student's life. It is where they are taught basic skills and critical thinking. As an instructor at the college and graduate level, I have found that students need personalized learning, and AI can provide that. I have always said that it's not that the student is capable of understanding something, thinking, or any complicated subject; it's the instructor's inability to teach them.

I wonder how many teachers have come to terms with that lack of ability and have written off students as incapable of learning something. It's a terrible thing to do to anyone, crushing their ability to learn and to make a future for themselves.

A Pew Research Center cited a poll done in the fall of 2023 that found most teachers are not sure about or see problems with using artificial intelligence (AI) tools in K–12 education. This is even though some teachers are using AI tools in their work.

About one in four public K–12 teachers say using AI tools in the classroom does more harm than good. Only 6% say it does more harm than good, while 32% say there is about the same amount of good and bad. Thirty-five percent say they are not sure.

AI can provide programs that act as digital mentors and provide personalized learning experiences. These programs analyze the student’s responses to lessons and then help them master those concepts at their own speed. Does anyone remember the programmed learning books from the 1960s and 1970s? They offered feedback in a manner that was both helpful and not derogatory. The element of emotion was removed, and that's an important step. AI does that, too.

These programs can also create much more engaging lessons, make learning fun, and make learning a task that students eagerly return to rather than shun away from. Homework can now be less intimidating.

Learning can even be incorporated into game-based programs that encourage students to strive for success. Students who have special needs, such as those with dyslexia, can have programs tailored to their individual needs, and non-English speakers can have specific translation programs that will be helpful.

The other day, I spoke to a primary school teacher who told me she was teaching her students earth science by having a garden near the school. When she talked about the environment and wildflowers, the children were astonished. To them, wildflowers didn't exist because flowers or plants only grew in pots. The children all live in inner-city communities, and when she took them to the beach, they found it startling that the ocean existed. AI can do much of what this teacher is trying, in a small way, to do: broaden the student's world knowledge.

For teachers, AI can provide administrative assistance in terms of lesson plans tailored to the students' needs, an analysis of grading and individual efforts or areas of improvement, and other guidance. Teachers are, therefore, relieved of some of the more mundane administrative work and can be much more creative and active with their students, along with these AI-provided guidelines.

As with everything, all is not rosy in the world of AI and there are serious privacy and security concerns. One concern of the National Education Policy Center is that student data can be compromised and exploited in some way. Remember, if it's on the Internet, it's potentially worldwide.

Another issue is cost, and these programs can be expensive, which would mean that only high-income areas could access them, and underprivileged children would be denied. This is what we call the digital divide. How many of those children with promise, given the proper tools, could have gone on to inspire the future, innovate technology, and become who they can become if not stymied by a lack of access?

Yet another concern is excessive screen time and the lack of independent critical thinking and problem-solving by students when they use AI. Screentime may affect a student's ability to develop appropriate social skills and engage in vital physical activity.

As always, we must be cognizant that no program is without problems, bias, and unseen mistakes. The training of these programs, how they are trained, and the population that was used for the training are of utmost importance. And not all AI programs are what they are touted to be.

Will AI totally replace teachers? That is a concern for many, but just as household appliances, never replaced our mother or our interactions with parents on an emotional basis, digital technology will not replace teachers totally. But this type of technology can bring education into the homes of children in areas where there may be inadequate schools available nearby and save them from endless hours on bus rides or the dangers inherent therein.

AI is here, and we must learn to evaluate it and secure it carefully to prevent it from causing harm to anyone or our society.

.