Roney Nascimento, an IB Diploma Mathematics Teacher and AI solutions developer describes his response to this insight.
“Teachers weren’t avoiding AI because they didn’t understand it. They were avoiding it because they didn’t trust it or trust themselves to use it responsibly. They were skeptical and wary.”
Gain insights for your professional learning design both in and out of technology adoption.
Read Roney’s ASCD article here.
Find Roney’s book, “Generative AI for Teachers: A Practical Guide to Educational Technology” here.
Subscribe to the Steve Barkley Ponders Out Loud podcast on iTunes or visit BarkleyPD.com to find new episodes!
Podcast Transcript:
[00:00:00.340] – Steve [Intro]
Welcome to the Steve Barkley Ponders Out Loud podcast. As instructional coaches and school leaders, you are tasked with guiding continuous teacher growth to create student success. This podcast looks to support you with strategies from experienced guests, as well as my own insights that I’ve gathered over the years. I’m thrilled to have you with us. Thanks for tuning in.
[00:00:29.960] – Steve
Building a Teacher’s Trust for AI Adoption. Joining our podcast today is Roney Nascimento, a AI solutions developer and IB diploma mathematics teacher at an international school in São Paulo, Brazil, and the author of, “Generative AI for Teachers: A Practical Guide to Educational Technology.” Ronnie recently designed and led an AI literacy program for his school staff, and he published his reflections in an ASCD article titled, “What I Learned from Training 50 Teachers in AI Literacy.” Knowing that many of the instructional coaches and principals listening to this podcast are planning similar learning opportunities for teachers, I was anxious to have him share his insights. Welcome, Roney.
[00:01:27.980] – Roney
Thank you so much for having me here. It’s a great opportunity. And let’s talk about this amazing AI introduction course that we started developing in some schools here in Brazil.
[00:01:43.550] – Steve
Talk to us about the task that was given to you. What PD were they asking you to provide to the teachers concerning AI?
[00:01:55.000] – Roney
So the task at the school that I was working with was straightforward on paper, but difficult in practice. It’s like the school and the entire faculty, roughly 50 teachers across every subject in year group, to develop genuine AI literacy. So genuine AI literacy, not a quick demonstration of ChatGPT, a deep working understanding of what AI can and cannot do in education. So I was given the brief because I sit at the intersection of three roles at school. So I teach IB higher-level mathematics in the classroom and I develop AI solutions for school, as you just mentioned. So that’s the idea of beginning of the task.
[00:02:43.360] – Steve
And when I read the piece that you wrote, you described that you took your common training approach when you got started, but you discovered that that wasn’t going to work. So talk about how you started and what you discovered.
[00:03:02.210] – Roney
Yeah, okay, that’s a great question. The big mistake I made at the start was assuming teachers need technical training. I built tutorials on Prompt Engineer. I showed impressive outputs and I got polite nods and zero implementation. It took me 3 weeks to see the real problem. Teachers were not avoiding AI because they did not understand it. They were avoiding because they did not trust it. And they did not trust themselves to use it responsibly. The shift came when I stopped preparing them to use AI and simply asked them to use it. To sit down, type a prompt, get a bad result, and look at it together. The moment they produce something imperfect with their own hands, the whole conversation changed. It stopped being, what can AI do? And became, why did I get this wrong for my student? And that’s a conversation a teacher already knows how to have. And I think so that was a shift moment on this whole training of 3 weeks.
[00:04:20.800] – Steve
I’ll tell you, Roney, when I read your piece, and you described that problem, first of all, it made total sense with the AI and what everybody’s trying to figure out now. But I think deeper, it jumped to me that I think that’s pretty common for people working in professional development with teachers, when we come to them with a new program and we run through a very short presentation cycle of this is how it works and go do it. And when they don’t go do it, we think that they’re opposed to it. But I think it’s probably much less being opposed to it than it is fear. I thought that word fear really jumped out at me when you used it. Fear that I don’t really know how to go about doing this.
[00:05:23.570] – Roney
Yes. And that happens a lot. Imagine that’s a new— it’s a new moment in the whole educational system if you think about starting to use generative AI. So for all the teachers that— and it’s true, for example, in Brazil here, I know a lot of teachers who’s not using AI, for example. They say that they’re not going to use…They don’t trust AI, so they try to avoid it. And I say, okay, I understand. It’s not mandatory for you to use, but I think you should try. So one example is when I talk about what AI can build for us. For the training that we build for the school, we go for 6 modules. Can I talk about modules?
[00:06:11.630] – Steve
Yeah, I want to, because when I read your piece, and by the way, listeners, I’ll put the link to Roney’s article as well as the link to his book in the podcast lead-in – make it easy for you to find. But he, in the article, he outlines the 6 modules, and I thought each of them were powerful enough that I’m gonna ask him to take a moment or 2 on each one. So Roney, I’ll name the title, because I thought you gave great titles to it in your article. I’ll name the title, and then you describe what happens happened in that module. But you called the first module, “Start with Ethics, Not Tools.”
[00:06:53.660] – Roney
Yeah, that’s because most AI training begins with look what this tool can do. Wow, that’s great. But I started with here are the downsides. So data privacy, algorithm bias, academic integrity, environmental impact. It’s a really big issue. So teachers did not leave excited. They left totally concerned. I was imagining that they would not come to the next session because they left, “okay, why am I doing this? I’m going to impact the environment? I’m going to go against the ethics? Oh, that’s a bias, that’s a problem.” And that was like, that’s how you need to feel about AI, generative AI in the first time, right? If I open this with enthusiasm and say, okay, let’s become a fan of AI, that could be weird. I need the teachers to know here’s the…
[00:08:03.010] – Steve
There’s an honesty there, huh? Let’s not cover up what there is to be afraid of. You put it right out there.
[00:08:11.580] – Roney
Exactly.
[00:08:12.730] – Steve
So the second module you titled, “Let Them Fail Safely.”
[00:08:18.000] – Roney
Yeah. So I give the teachers a low-stakes challenge. So plan a lesson on a topic you have never taught using AI. So no guidance, no templates, just try. A math teacher, decades of experience, types make a lesson on fractions. And got a plan to teach at 16 years old for her class of 11 years old. That’s pretty bad, right? So she laughed and said, okay, this is actually why it needs me. I said yeah, but that’s the beginning. So that one sentence did more for the program than any model I wrote. So she reframed AI from a replacement to a collaborator. And she said it herself, it did not tell her, that is the whole point of jumping in. So imagine this situation. So she tried to use AI, she got a worst result possible for 11 years old. I say, okay, so it’s not perfect. It’s not like, it’s not getting my job.
[00:09:32.800] – Roney
I can use it and I can teach AI how to use AI or how to work with me. That’s changed everything in that moment.
[00:09:41.560] – Steve
So she discovered two things: that AI needed her, so that built the security. But then I guess she also discovered she can teach AI. She could fix AI how to do what it was that she wanted it to do.
[00:09:58.180] – Roney
Yeah, because AI will be still there. It’s still here. It’s between us. It’s static. We’re not moving. If you don’t do something, nothing’s gonna happen.
[00:10:12.000] – Steve
So your third one was, and I thought this is powerful, “Co-create, don’t consume.”
[00:10:20.310] – Roney
Okay, so that was a good module. So instead of showing teachers pre-made AI tools, I asked them to build their own. So small custom chatbots tailored to their subjects using platforms like Magic School. We used to use a lot Magic School. They have a lot of options for creating this kind of tools. So a history teacher built a Socratic dialog bot that never gives direct answer and only asks probing question. That was for me amazing. That was a great idea because everyone said, “Okay, so we can do this?” And I said, “Yes, that’s the great part of the system with AI.” You build new ideas, new systems, and science teacher built a misconception detector, which was pretty good. So in that moment, teachers stopped seeing AI as something that happens to them. So they start seeing it as something they control. So if you can control something, you can do whatever you want to do with that, right? And once teachers felt they are in control, resistance drops. That was a great thing about Module 3.
[00:11:42.690] – Steve
So Module 4, Address the integrity elephant.”
[00:11:47.390] – Roney
Yeah, okay. So imagine this, okay. So every AI training hits the same wall, but students will just cheat. I did not dismiss this. I reframe it in a way like, students could already cheat with calculators, the internet, older siblings. So the question is not how to stop cheating, it’s how we are actually assessing. That was the big elephant in the room because of this question. So an English teacher, I had few as a skeptic, resigned her literacy analysis task. What’s happened? So instead of asking pupils to write an essay about a novel, she asked them to critique an AI-generated essay about that novel. So they need to identify where the AI reads tone, context, character motivation, and in that moment, the AI becomes part of the assessment, not a shortcut around it. So for this teacher particularly, I cannot say his name, but I think that was one teacher and a other teacher. So she said something to me after and said, okay, now I can see where that’s going. I can imagine all the students use AI. I’m not going to try to stop the students from using AI. I’m going to use as a tool to help me make them understand better novels. And I said, “Okay. Great.”
[00:13:19.270] – Steve
The thing that I discovered some people writing about that integrity issue really clicked for me – that what we really have to be doing, and I think this works into what you said, this teacher design, what we really have to be doing is assessing the process rather than the product. The degree to which the student has to show how they got to where they are, now I know the learning took place versus if all they had to do was show me the finished piece, then that’s where we can get away with the student, what we might call cheating.
[00:14:09.040] – Roney
Exactly. That’s perfect.
[00:14:12.460] – Steve
And the fifth module you did, you titled, “Show Real Data.”
[00:14:18.640] – Roney
Yeah. Okay. So halfway through the program, I shared preliminary data from my doctoral research at the University of São Paulo where I’m studying how IB diploma pupils actually experience AI assistant tools. And the headline is that AI reduces cognitive loads on low-stakes tasks, which everyone expected. We have already a lot of studies coming from, I think, Stanford, if I’m not mistaken. So the more interesting find is that on high-stakes tasks, or where we’re actually relying on learning, teachers responded to the honest. Several told me it was the first time anyone had shown them real evidence of AI limitations along its strengths. And that changed how they thought about using it. One thing that we saw happened a lot in this part is that we have— you can work with real data with a lot of different AI tools and you can build your AI tool to work with real data. I think that in that moment, the teachers understand and know that if the people just use AI to do all the work, everything, they’re gonna lose the interesting part of using AI, which is learning better and finding your own way of learning.
[00:15:35.290] – Roney
And I think that’s something really important for us to think about. If you want to use AI, that’s amazing. That’s a great tool. But like, how are we using AI to make sure that the pupils are really learning everything and not just passing through this moment without learning, without having the thinking, critical thinking. That’s important.
[00:16:01.140] – Steve
So that really connects back to that piece that I had read then, that the teacher having the students report on the process becomes important so that I know – I’m thinking in some spots it’s even asking the students how they used AI. You know, almost having the student share with the teacher, “When I got this result from AI, I asked it to do this.” And if the student can explain that to me, that’ll tell me what the student’s thinking is. It’s really interesting – I think I’m finding as I use AI to write, like writing a blog, I think I actually end up doing more thinking because of my work with AI, because I have the time and the flexibility to look at it from several different directions, where if I was sitting down and writing it on my own, I wouldn’t be able to dedicate that time. So the critical thinking has actually increased. Which if you didn’t talk to me and all you did was look at the outcome, you wouldn’t know what I went through and the thinking that I did to get there.
[00:17:19.730] – Roney
Exactly. And I think this point of this process in this way that you just mentioned is like you have a digital version of yourself inside some AI tools because you put all your information, like imagine this – and they know how you feel, they know what you’re doing, they know exactly what’s happening now. So for example, before we start this meeting, I didn’t find my folder and I sent a message to Claude and say, “Okay, I got here, my car broke, I need to talk to Steve now.” And it says, “Oh yes, you’re right on time, here’s the paper.”[laughter] And I say, “I’m a little off time here, but it’s okay.” I say, “Oh, I wish everything’s okay with the car.” I say, “okay, don’t worry, bye, we talk later.” [laughter]
[00:18:09.812] – Steve
[laughter]
[00:18:10.100] – Roney
And he sent me the prompt. I think that’s amazing. That’s the interesting part of using AI in this way because the AI works with you and that’s amazing. I love this part.
[00:18:21.180] – Steve
The last module you did was called, “Build a Community.”
[00:18:25.210] – Roney
Yeah. Okay. So the final module was not about AI at all. It was about sustainability. I help teachers form small learning communities, 3 or 4 colleagues who would continue to experiment together. So 6 months on, those communities are still active. In fact, now they are more active than ever because of these new models of Anthropic. Teachers who also avoid AI now share prompt libraries in WhatsApp groups.
[00:18:57.110] – Roney
Here’s what surprised me. The single best predictor of whether a teacher was using AI 6 months later was not how good they were in Module 2. It was whether they had 2 or 3 colleagues in a group sharing prompts. The technology did not sustain adoption. The trust between the peers did. So the trust changed everything.
[00:19:20.540] – Steve
Yeah.
[00:19:21.420] – Roney
And I can see because like, imagine this situation. All the teachers between the small communities and between themselves, they send messages, they send programs, they send websites, they build something, they said, “Hey Roney, take a look on this.” I said, “Whoa, that’s great! You just built that?” “Yes, and it’s working.” I said, “That’s amazing. You used the AI in the best way. Thanks for sharing.” And now, in this moment, if we have a module 7, which we don’t have, I would say module 7 would be sharing is caring. Because all the teachers are sharing their finds, their new tools, their work between themselves, sometimes with the pupils. So at the beginning, the pupils were much more smart in using AI in the school. But as we had a lot more experience, now that changed. Now I can see the teachers using AI in the best way in the school. That’s really great. That’s amazing.
[00:20:26.310] – Steve
That’s cool. Well, you captured a whole lot of my work there because my work is driven by the phrase “teaching is a team sport.” And so what you discovered in that community piece is a big reinforcer on that. I wanna put a little twist in here before we wrap up. One of the issues that’s getting a big amount of discussion for sure in the States right now, but even broader is the amount of screen time kids have. And initially, people were looking at how much time kids are on their phones or computers at home, but now there’s kind of a pushback saying as to how much time they’re on it at school. And I wonder, in the work that you’re doing and the research you’re doing, if you’ve got some thoughts about how educators ought to be looking at that issue of kids on screen?
[00:21:36.570] – Roney
Okay, that’s a really, really interesting question because in this moment we live in the screen almost all the time. I think that it’s really hard to see people out of screens – inside families and all of that. It’s like, I felt that and I can see that in all my pupils. Okay, let’s say that screen time is not a problem of technology. It’s a problem of attention. And attention and it’s a skill that has to be taught.
[00:22:08.910] – Steve
You got it. And it is something we need to teach.
[00:22:12.770] – Roney
Exactly.
[00:22:14.150] – Steve
Yeah.
[00:22:14.570] – Roney
And I think that if you go this direction and thinking this way, we probably can find better ways to use the screen time when we need and as a better partner, not just using screen time all the time without having any kind of plan or attention for the details, because that changes everything. And the attention, I think that the attention is the real problem behind all the screen time here.
[00:22:46.310] – Steve
Yeah, it’s interesting. I was just reading about a school that they passed a guideline, no technology during recess or free time. And that made a ton of sense to me, you know. If you’ve gotta be on the screen to do your work, then, you know, if we got a 15-minute break, the idea that you’re gonna sit and scroll and find a YouTube video to watch probably wasn’t the way to spend your break time. It may start with rules, but you really come back to it is that we’ve got to build the learning in.
[00:23:26.950] – Roney
So that’s— we know that’s going to change forever. And I don’t think— I don’t think what you believe, Steve, in this case. But if we see the education for now from 6 years from now, 5 years, let’s say, what do you think is going to happen with the screen time? I don’t know if I can ask you that because it just came into my mind.
[00:23:46.970] – Steve
Yeah. I don’t think we know. So what we have to do is—I use the phrase, way back prior to this whole issue of it’s about learning how to learn. And learning to manage your own screen time, may be part of learning how to learn. Now if that reflects back on schools, it means we got to be able to give kids the options. If the only way you can do the things that have been assigned and tasked to do is screen time, then we need to be a little bit more careful. But looking at how we engage students – I just heard one teacher who was being interviewed on this, and she added one interesting thing that she was elementary, but she started giving kids tasks to do, but one computer for two kids. And so what it did is it automatically built a social interaction, human-to-human piece, even though the students were on the computer in order to do the research and pull the material they needed. But by simply going to one computer with two kids, she added another piece.
[00:25:07.060] – Roney
Wow, that’s interesting.
[00:25:08.650] – Steve
Yeah, it’s going to be tons of things for us to be exploring here.
[00:25:12.700] – Roney
Yeah. Okay. I thought about that. That’s something you just put in my mind, the situation. So human to human and to machine.
[00:25:22.290] – Steve
Yeah.
[00:25:22.690] – Roney
Okay. That’s interesting. I think that’s good. That’s something.
[00:25:26.880] – Steve
So what do you think? As we wrap up here, I’ll remind people that the links to the book and the article are in the podcast lead-in. And folks, I really recommend the article to you even if you’re not doing PD for AI. I think that Roney laid out a bunch of items there that just jumped out at me as being natural to building a teacher learning environment. But wrap us up here, Roney. What do you think might be— you got a guess of what some of the most exciting things might be for learners and schools with where AI might take us?
[00:26:10.600] – Roney
So, okay, I can see a group of people from China came this month and they show me a lot of things that China using with AI. And I say, okay, so what’s going to happen in the future? I imagine, all the teachers and students will use AI in the best way. Focus on the learning, not on the building something for show the teacher and the teacher gets the answer to show the coordination. So that’s a best way of learning and learn – that’s the best thing that I learned today from you, Steve – learn how to learn using AI tools. I think that if all the people around teachers and students, learn how to learn with AI, I think that the future can be much better, much greater for everyone. Because we cannot avoid AI at all, we know. So let’s learn how to learn using AI tools. And that’s one part from Steve, one part from me. And mix the ideas and learn how to learn using the AIs. I think that that’s a massive thing and that’s something you given in my mind today. So thank you so much for that, Steve.
[00:27:36.810] – Steve
Well, thank you. I appreciate that. I’ll continue to follow. I got to guess you’re going to be writing more articles that both go with your research, but also with your finding with students in the school. Much appreciated, Roney.
[00:28:00.330] – Roney
Much appreciated also, Steve. Thank you so much. That was really, really great talking to you, learning from you – everything. I wish we would have more time one day. And if we have more time, I would love to talk to you more about AI in the future.
[00:28:18.940] – Steve
We’ll do it again. Take care.
[00:28:21.430] – Roney
Thanks so much. Take care.
[00:28:24.870] – Steve
Thanks for listening, folks. I’d love to hear what you’re pondering. You can find me on Twitter or LinkedIn at Steve Barkley, or send me your questions and find my videos and blogs at barkleypd.com.
