AI & My Business – Find the middle way to get the balance for AI in education

And just how many AI searches does a single sunflower offset in carbon emissions? 

As attitudes to AI in education mature, we need to move away from a binary attitude of threat versus opportunity, and move authenticity, sustainability and equity to the fore.

That’s according to Rita Bateson, Chief Education Officer and Co-Founder of Eblana Learning, who is witnessing a shift in perspectives around AI among teachers and students alike. 

New mindsets for AI in education

“Back in late 2022, generative AI became accessible and for the first year or two the talk in education was dominated by the use of tools and there was a lot of fear about the impact on education and particularly on cheating in assessments,” says Bateson. “Managing that put a focus on stopping students from using it, and on getting AI literacy up among students and teachers.”

Now though, we are entering a more reflective era, she notes, and this is changing the mindset.

“We are seeing a shift now where educators are focusing on critical discernment, on developing a better understanding of what AI produces, the mediocrity or flatness of it, and on learning how to evaluate hallucinations and biases,” she says. 

Bateson and her colleagues at Eblana work with International Baccalaureate schools across Europe, the Middle East and Asia. They help them to develop the governance, skills and curricula to harness the opportunities that AI offers for students while minimising the downsides, such as cognitive offloading and a lack of connection. 

“Private international schools have a mandate to innovate, and we can all learn from that,” she says. “We help these schools to become AI-empowered, and we share what we learn from that with the wider education community.” 

The middle way

Bateson started her own journey with AI in education as a skeptic. 

“I was a school head and a curriculum writer, and AI felt very much like an existential risk for educational values such as understanding knowledge,” she recalls. 

But when Bateson met International Baccalaureate educator Rachel Bodily, they had a meeting of minds and Eblana was born. 

“Rachel was an AI cheerleader, I was an AI skeptic,” says Dublin-based Bateson. “Between us we duked it out to figure out the positives, such as AI being amazing for efficiency and productivity, and the horrific risks, such as the threat to cognition and creative thinking and the avalanche of mediocre ‘AI slop’ we are all subjected to. We found we could have a more nuanced approach about the deeper issues at play, and we developed an AI framework for schools to accelerate the amazing use cases while minimising the dangerous elements.”  

Working with schools around the world, Bateson has become aware of a wider binary discourse that needs to converge. 

“A lot of discussion, particularly online, falls into two camps – either you’re in the pro-AI camp, with talk of AI literacy and efficiency, or you are in the ethical camp, with talk of doom and gloom and brain rot,” she says. 

Much like the Eblana co-founders, those discussions need to find a middle path, she notes: “We need to balance the ethics and the considerations with the power and potential of the technology.”

Share the knowledge

The ethical questions that need addressing include equity, and Bateson is conscious of an emerging imbalance in access to AI in education.

“I think it’s important that we don’t have another digital divide, that just because international schools have the finances to experiment with AI and learn how to use it, that inequities in AI don’t get perpetuated,” she says. 

She recently published a book – International School Leaders’ Guide to AI – which includes audits, ready‐to‐use policy templates and stakeholder frameworks to help everyone in education get the most from the technology in a responsible, safe and empowered way.   

Bateson is also keen to share the learnings widely in Ireland and Eblana has recently joined the Learnovate network as a member. The company encountered Learnovate when looking for a ‘critical friend’ to help develop AI curricula for schools, says Bateson. 

“Learnovate were really proactive,” she says. “They helped us to shape the research questions, and then really tried to rip the curriculum apart and look for gaps, to make sure the end result was rigorous. And now through the Learnovate network, we are looking to contribute to the Irish educational system and community.”   

Stop the slop and make the connection

One of Bateson’s key concerns for everyone is the rising online tide of mediocre or misleading AI ‘slop’ that lacks authenticity and genuine connection.  

“It’s creeping in everywhere,” she says. “And in education and beyond we need to develop the awareness of it and the confidence to not get drawn into it.”

Encouragingly, she sees students naturally pushing back on the blandness of resources that are shaped predominantly by AI, with many demanding a more human connection. 

“It’s interesting, many students get frustrated by the flattened examples and similar-looking resources from class to class and they don’t want AI-driven feedback,” she says. “Instead they want a person to read and comment on their work, they value the teacher’s contributions.”  

Count the carbon cost

One of Bateson’s main drivers now is the ethics of sustainability and responsible energy use in AI. 

“We have a real energy crisis, and education has a part to play,” she says, “We want schools to understand the impact of non-conscious consumption or reckless adoption, of burning carbon for the sake of cat memes and deep fake videos.”

Bateson conveys the impact with eye-opening comparisons. 

“People sit up when I say that the energy needed to process 10 prompts is the same as the energy you expend running for 17 minutes, or how an AI-generated image requires the same energy as charging your phone three times,” she says. “Or that one 30-second deep-fake video uses the same energy it would take to play all of the episodes of Friends from the whole series twice.”

Conscious of their own footprint, Eblana tracks its AI use and offsets the carbon cost by paying for the planting of native trees and sunflowers, says Batson. “If you plant a sunflower, you can probably offset around 300 prompts or so, if it grows to full capacity,” she explains. 

Heed the curse of convenience

Bateson sees parallels between AI and our previous experiences with plastics and ultra-processed foods, where convenience can lead to overuse and problems. 

“We saw this with plastics, which can be wonderfully useful materials, but how they were overused and thrown away too quickly, and now they contribute to environmental issues. Also with ultra-processed foods, which are cheap and convenient but again eating too much of them can lead to problems,” she says. 

“If we overuse AI in forms that present ultra-processed information, we risk issues of energy use, cognitive offloading and a loss of engagement and creativity. In education we need to ensure that every school is empowered to use AI in a way that enables and nourishes minds without the damage.”