Listen to Renée read this article aloud.

It feels like I’m standing on a wide open beach. As I look out at the ocean I can see a big wave approaching, but it hasn’t yet hit the shore. That’s my answer when I’m asked about the impact of AI on students with dyslexia.

The AI conversation is well underway — that off-shore wave is clearly visible, generating strength, power, and potential. But when it comes to the ultimate effect it will have on the way students with dyslexia learn and process information, we are still in the very early stages.

Like most school leaders these days — particularly those at schools like Carroll, where the possibilities of AI are so meaningful, so significant to students — I’m considering all the angles. 

Here are my initial thoughts.  

“Both / And”

With the help of Google’s Read&Write software — including predictive writing, voice-to-text, and text-to-voice features — AI is already supporting our students’ ability to comprehend text and formulate and express their ideas.

But as I watch Carroll students benefit from this technology, I fit squarely in the “both / and” camp: I am both curious and cautious; excited and reserved. I see both AI’s possibilities and opportunities and also its liabilities. 

I am reminded of our transition to a 1:1 computer program, which launched in the early 2000s, when I was a Carroll teacher. Conversations and questions among us were plentiful. The multisensory experience of handwriting is so important for students — will that be lost with typing? Will they ever learn to spell if they rely on spell-check?

What we were really pondering is: How do we maintain the instruction we know works well while incorporating this new piece of technology in a meaningful way? 

The stakes may be higher, but today’s conversation around AI feels very similar. We can be excited by its potential for our students and alert to its shortcomings at the same time. 

AI as a classroom conversation-starter

ChatGPT is not widely available to Carroll students. Are they using it at home? Probably. Rather than discipline them, we can start some pretty productive discussions by asking them:

• Why are you using it? 

• How do you feel it helps you? 

• Do you notice any disadvantages?

We teach into that conversation. And while our students aren’t using ChatGPT in the classroom for assignments, our educators actively model its healthy use. For example, they encourage students to experiment with the questions they pose and together, they analyze how factual the results are.

Our values around teaching and learning are our North Star

Teaching students to think critically and independently is — and has always been — central to our work. Conversations around AI are just another opportunity to reinforce these values. When it comes to generative AI, we ask them to consider: 

• What kinds of questions can you build to generate a more factual answer? 

• What perspective is this answer coming from? 

• What perspective(s) is missing?

• What does it mean to read a primary source versus the internet? 

When AI feels big and uncertain, I remind myself that our teaching and learning values remain constant and unchanging.

Leveraging technology in support of the academic goal

The same way we teach our kids to be critical consumers of content, we will continue to teach them spelling, grammar, phonics, sentence structure, and how to organize an essay. Technology doesn’t replace these traditional instructional methods, but it can be leveraged to help students successfully meet their goals. If the goal is to write a five-paragraph persuasive essay, speech-to-text can be a tremendous tool, allowing them to more easily express their ideas and freeing them up to focus on higher-level thinking. The task of proofreading remains, but the technology helps them to clear a processing hurdle in order to achieve the end result.

With achievement comes confidence and empowerment. And with confidence, even greater achievement is possible.

Protecting student data is our primary concern

Given the age requirements for accessing AI tools (not rated for individual use for students under 13), we’re exploring high-quality tools to support educators’ practice. We’re also looking at how we can use AI to expedite the synthesis of student learning data. If teachers can more quickly and effectively absorb a child’s learning profile, they have more time to spend customizing interventions.

Right now, protecting student data and identity is our utmost priority. We take seriously the ethics around what we’re sharing with AI programs and how. We are grateful to Dr. Fernanda Viégas P ’24, who, with dual appointments at Harvard and Google, has been helping us think through safe and secure ways to harness AI for our learners.

Helping teachers navigate the AI wave

Carroll educators are just beginning to build their AI understanding and ease of use. Those with dyslexia are particularly eager to try out new tools. Some are using AI to generate word lists for their OG focus areas or for connected text. Others, as mentioned, are demonstrating the use of ChatGPT in class to start productive conversations around its advantages and limitations.

Teaching our teachers how to leverage AI in the classroom is just as important as guiding our students. We thank our Director of IT, Bishop Levesque, and Director of Academic Technology, Erin Jacobsen, who are leading this work. We can’t live in fear of change, whether that’s the advent of laptops, AI, or whatever comes next. Supporting our teachers to embrace innovation — to benefit students — is key.

Embracing change, but staying the course

As I stand on the beach, watching and waiting for the AI wave, do I feel a little uncertain about exactly how it will land? Of course. But I’m confident that the skills we’re already teaching our students — reading, writing, executive functioning, critical thinking, problem solving, perseverance — will remain our guiding force. AI will be another essential skill we teach them to use, another tool for their toolbox. 

As Dr. Viégas points out, ignoring AI is not an option. “I don’t believe in telling students not to use this technology. The cat is out of the bag. Let’s not pretend it doesn’t exist,” she told the Born Curious podcast last spring. Instead, she adds, let’s ask: How can we use it to the maximum benefit of students?

For a helpful introduction to AI, check out the Born Curious interview with Fernanda Viégas P ’24: “Artificial Intelligence: How It Works.”