Teacher retention is an issue that’s top of mind for many K-12 education leaders. In trying to sort though that tangled staffing issue, school and district leaders may be able to turn to a surprising resource—artificial intelligence—for ideas on solutions.
That was the message offered by a presenter at this week’s ISTELive26 + ASCD Annual Conference, who focused on the technology’s potential beyond supporting educators and students in the classroom.
Administrators don’t have to build a custom AI tool to leverage the technology’s power, nor should they just encourage teachers to use AI to lighten their workloads, said Nick Schiner, the director of peer-to-peer professional learning at Digital Promise.
Instead, AI can help school and district leaders work through the specific challenges in their buildings that are leading to teacher turnover, said Schiner, who has been working with Google Education and school leaders on developing best practices for querying AI to synthesize research and identify weaknesses in their planned solutions.
“How can we do root-cause analysis of this?” Schiner said. “It’s about using [AI] as a tool for thinking, as a tool for having something to respond to when you don’t have another person that you can work with in that moment.”
Schiner presented on the topic at the massive ed-tech show, playing out in Orlando, Fla., this week.
Here are five tips Schiner offered on how to leverage AI to get meaningful feedback and ideas on how to improve teacher retention—and potentially, to help solve other persistent problems bedeviling school and district leaders.
1. Give AI tools ample context
The more detailed information provided to an AI chatbot, the more tailored the outputs will be.
Schiner calls this “building your context block,” and it should include details such as the prompter’s job title, the location of their school or district, and how many students it serves.
Public data, including information on student academic performance, can also be fed the chatbot, along with information on any major initiatives the school or district is undertaking, said Schiner.
“Not even necessarily initiatives related to teacher retention,” he said. Tell the chatbot: “These are some of the initiatives that we have going on. Because what we’ll find a lot of times is [the AI] will say like, ‘Hey, you’ve got a lot of initiatives going on in your building.’ This actually might be one of the things that is contributing to teacher burnout and teacher turnover in the school.’”
It’s important to not give AI any personally identifiable or private data.
2. Tell the AI tool, ‘do not offer solutions’
AI chatbots are often eager to solve users’ problems—but that’s not what school and district leaders should be asking the tech to do, at least not yet. As they begin their work, school and district leaders should be asking whatever AI tool they are using to help synthesize research, brainstorm ideas, and spot gaps in their thinking.
The best way to counteract this, said Schiner, is to specifically tell the chatbot in each prompt not to offer solutions.
“It’s a way to make sure that the model is doing as little of that solutioning, upfront,” he said. “It’s both a way to get better results from the prompt, but also to remind the end user, the school-based administrator in this case, that they are not turning over solutioning to the tool.”
3. Prompt AI to play devil’s advocate
Asking AI tools to do this can help school and district leaders test their ideas and plans and get critical feedback that their direct reports might be hesitant to provide. If school leaders are only getting positive feedback on their plans and initiatives, it might be time to ask a chatbot to make counter-arguments and tell them about possible flaws in their proposals.
“It’s a really great exercise to go through when you’re not sure that someone [on your team] is going to be like, ‘Is this a good idea?’”
While AI can help play this role, it’s ideal to also have school faculty and staff who are ready to offer critical feedback on ideas, Schiner said. If school and district leaders don’t feel like they’re getting honest feedback from their teams, Schiner said it might be worth asking an AI platform for ideas on why that is.
4. Ask: what—or who—am I missing?
We all have blind spots (AI models do, too) but one way to identify things we’re missing is to ask AI.
For example, Schiner has worked with principals who realized through their brainstorming with AI that the mentorship initiatives in their schools focused on new and struggling teachers.
“As they dug into that problem of practice, going back and forth with whatever AI tool they were choosing to use, they were getting things back that were making them think about, how are we mentoring our strongest teachers?” He said. The exercise forced principals to consider if asking their highest-performing teachers to always step into mentorship roles was contributing to their burnout, Schiner said.
“How are we making sure that our teachers who are successful, whom we often lean on for those mentorship roles, how are we making sure that they are getting support?” he said.
5. Practice what you preach on AI literacy
It’s vital that education leaders don’t offload their decisionmaking and critical thinking to an AI tool. This is where they should practice some of the same AI literacy skills that teachers ask students to adhere to, said Schiner.
When considering responses that come out of AI tools, administrators should ask themselves several questions, such as: Does this response sound accurate? Does it reflect their schools’ context or situation—or is it too generic? What assumptions is the AI making? Whose experience is being centered, and whose perspective might be missing?
“That’s why we prompt and then we prompt again, and then we pressure-test and we bring other humans into the loop and then we go back to it again,” Schiner said. “Because we know that one person working with one AI tool is still just one person working. [It] can oftentimes seem like another person, but at the end of the day, they are doing that on their own.”
