5 questions schools should be asking about AI

(TND) — An expert in how technology affects education has come up with questions schools should be asking before taking the leap on “shiny new, very promising” generative artificial intelligence tools.

“There’s been a lot of promises about technology and education, and oftentimes those have failed educators,” George Veletsianos, a professor of learning technologies at the University of Minnesota, said Tuesday.

Veletsianos said schools risk harming student learning if they don’t ask the right questions.

The five questions that he said school officials should be asking before buying new AI-powered tools:

Which educational problem does the product solve?

Is there evidence that a product works?

Did educators and students help develop the product?

What educational beliefs shape this product?

Does the product level the playing field?

Veletsianos went into detail on each of the questions in a new article for The Conversation.

There’s a spectrum of enthusiasm within the educator ranks about generative AI, he said.

Some folks are really excited about it. Khan Academy’s founder said in a TED Talk last year that AI offers the potential for “the biggest positive transformation that education has ever seen.”

Veletsianos said other educators are concerned about what AI might mean for teaching and assessment practices. Some are worried about loss of skills.

Veletsianos said schools should approach technology with caution.

“Perhaps cautious optimism, if you will,” he said.

A previous study pointed to the power of AI as a brainstorming tool for students, but not without its potential pitfalls.

And Stanford education scholars have found that concerns about students using AI to cheat seem to be overblown, at least so far.

Veletsianos said schools weighing the use of AI ought to be asking questions guided by past lessons, which have shown ongoing failures of technology to transform education.

He said he doesn’t think any technology by itself can transform education at large.

“But there’s room, of course, for improvement,” he said.

He wants school officials to understand what the technology is and what it does. Try it out themselves and see how it can help them address problems they might have in their classrooms.

There’s no need to replace existing practices that work well, he said.

Veletsianos said he also hopes developers involve real teachers and students. Otherwise, they risk building tools divorced from the realities of the classroom, he said.

And schools should make evidence-informed decisions. Schools should ask vendors to provide independent studies of their products, he said. Educators should be on the lookout for unsubstantiated claims.

The wrong tools could waste precious school funding. But they could also do actual harm, he said.

AI could introduce biases in our classrooms, he said.

And schools that buy an AI product might be inviting ways of teaching and learning that are inconsistent with their approaches, he said. For example, an AI model might take on the role of “expert” and just give students the answers when all educators want it to do is spark follow-up questions or collaboration.