As AI continues to improve, its applications for teaching and learning are expanding. In this week’s opportunity forecast we’re exploring AI learning partners (AILPs). Over the course of five sections you’ll learn what AILPs are, who uses them, and what the future for them might be.
The five sections are:
- Day 1: Introduction – what AILPs are and how they are different from traditional AI tools
- Day 2: Current Market – who the different user groups for AILPs are
- Day 3: Implementation Challenges – the tradeoffs between stakeholder groups you need to balance when implementing AILPs
- Day 4: Educational Value – the research into the uses and effectiveness of AILPs
- Day 5: Forecast – what are the pillars of AILP capabilities and what are they likely to evolve into in the future?
As you move through the sections you’ll do a series of reflection activities that build on each other as you think through the opportunities and challenges of these powerful new tools.
To get the conversation started, respond below with your thoughts on one of the following questions:
- When you think of an AI learning partner, what do you imagine it does, and how is that different from a search engine or chatbot?
- Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable?
- By the end of this lesson, what is one question you hope to answer about AI learning partners or their role in education?
The Learning Partners team:
~ Sut Teng Chong (Ice), Jason LeHuquet, Sam Quarterman, and John Rose
I would imagine that an AI learning partner is not only there to answer questions, but to also guide you to the answer of the assignment through questions and dialogue, rather than just giving you the answer. In terms of learning something difficult, most of the difficult topics I have learned involved complex thinking where there are multiple perspectives or higher level math. An AI learning partner might help with one on one instruction with structured and levelled questions, but I also think that it is harder for it to evaluate your written math work or point out misconceptions in a way that students remember in their long term memory. One question I have about AI learning partners is which ones exist and if there have been studies done on their viability in education, and which contexts it is most viable in.
Thanks for being the first to share your thoughts, Logan! I really like the distinction you made between an AI Learning Partner guiding learners through questions and dialogue rather than simply providing answers. That idea connects closely with one of the key design trade-offs we explore in the OER.. how much support is enough to help learners without taking away opportunities for productive struggle?
You also raised an interesting question about where AI Learning Partners are most viable in education. As you work through the later sections, I’m curious whether you’ll find that your answer depends more on the subject area, the age of the learners, or the way the AI is designed. We’d love to hear your feedback on this later in the week!
Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable?
For me, learning languages has always been difficult. I think an AI learning partner could help by providing conversation practice, pronunciation feedback, and vocabulary support. It could also provide immediate feedback and adapt practice activities to my learning needs, which would help build my confidence. However, it wouldn’t replace the authenticity of being immersed in the language or practicing with real people. While AI could be a valuable tool, I think real conversations and cultural experiences are still an important part of language learning.
Thanks for sharing, Jodee! I really appreciated your example of language learning. AI Learning Partners have the potential to provide immediate feedback, personalized practice, and opportunities to build confidence in ways that aren’t always possible in traditional settings.
Your point about authentic conversations is also an important one. It raises an interesting question about the role AI should play… should it aim to replace aspects of learning, or should it primarily prepare learners to engage more confidently with real people and real-world experiences?
As you explore the Educational Value section of our OER later this week, I’d be interested to hear your thoughts on this!
Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable?
There have been times when I have tried to learn a new song on the piano. I read the notes and slowly put the tune together. Then after a week of playing I will go to an instructor and they will say “see that note there, that is an F not a G” and I go, ah! how did I not notice that before!? I then have to spend time unlearning what I have already trained my brain to do. With an AI program listening in while I practice, it could pick up on wrong notes right away and correct me. In this type of scenario an AI instructor, could be helpful. Not to replace, but to suplement the real thing.
Fantastic example, Bryce! I love how you distinguished between AI as a supplement rather than a replacement for an instructor. I think that highlights one of the strengths of an AILP, being able to provide immediately, personalized feedback before habits happen that are difficult to unlearn for piano. However, at the same time, I think your example also raises an interesting design question. Should an AILP always step in immediately when it notices a mistake, or are there times when allowing a learner to self-discover an error might lead to deeper learning? I’d be interested to hear what you think after exploring the trade-offs and Educational Value sections later in the OER.
When I think of an AI learning partner, I imagine something much more collaborative than a chatbot or search engine. Rather than simply providing information or answering questions, an AI learning partner should understand where a learner is in their learning journey, adapt its support over time, ask questions that promote deeper thinking, and provide feedback that helps move learning forward. In that sense, the relationship is ongoing rather than transactional.
One question I’m hoping to explore is how AI learning partners can support authentic learning without replacing productive struggle. Some of my most meaningful learning has come from working through uncertainty, making mistakes, and reflecting on feedback. If AI provides answers too quickly, it risks reducing opportunities for thinking, processing, and reflection. Learning takes time, and I’m interested in how AI can scaffold learning while still leaving the cognitive work to the learner.
I’m also curious about who is building these learning partners and the assumptions embedded within them. Who we learn with, and who we learn from, profoundly influences human growth and development.
Hi Denise,
Thanks for adding to the conversation!
I appreciate the distinction you make between an AI learning partner being collaborative vs. transactional. Your point about supporting learning over time, rather than providing or answering questions captures one of the central ideas we aimed to explore in our OER.
I also think you make a great point about productive struggle. Finding the balance between providing enough support to keep learners moving forward while preserving the cognitive work of learning is one of the key design trade-offs we explore in both the Implementation Challenges and Educational Value sections. It’s also important to consider who is building these systems and the assumptions embedded within them. This raises broader concerns about whose knowledge, values, and perspectives shape AI learning.
By the end of this lesson, what is one question you hope to answer about AI learning partners or their role in education?
The main question I hope to answer is “What are AI Learning Partners” as this is the first time I’ve heard this term, and how it can support Principals and Vice Principals in elementary schools. I recent hosted a Pro-D with teachers and when it came to AI, I was shocked to learn that many teachers are hesitant to use AI (mainly because of ethics). I’m hoping AI learning partners can assist in creating a Pro-D session focussed on AI/AI learning partners and to assist teachers with being comfortable in using AI/AI learning partners. Looking forward to learning about AI Learning Partners!
Hi Jessy,
Thanks for sharing your perspective. I actually wasn’t surprised to hear that many of the teachers in your Pro-D session expressed hesitation toward Ai. In my role supporting EdTech in my school division, I’ve encountered many of the same concerns, particularly around ethics, privacy, academic integrity, and uncertainty where AI fits within teaching practice and pedagogy.
One thing our groups hopes partipipants notice as they work through the OER this week is that we intentionally distinguish AILP’s from AI tools that generate answers. Many of the examples focus on scaffolding thinking, providing feedback, encouraging reflection, and supporting the learning process rather than replacing it.
Looking forward to hearing what stands out to you. Do you think framing AI as a learning partner instead of an AI might change the perception teachers have in using it in their practice?
Hey Sam and team,
I think framing AI as a learning partner instead of an AI will definitely change the perception teachers have in using it in their practice. I used the final activity to create a ProD based on AILP’s! Will tweak them to fit my staff’s needs. A great OER, thanks for the learning, I’m walking away with exactly what I was hoping to walk away with from this OER!
Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable?
I think an AI partner would be great for learning languages. From past experience living overseas and trying to learn a language, I mostly relied on tutors (actual humans) who can answer my questions from having direct knowledge of the language and culture. An AI Learning Partner might be helpful for learning the structure of the language, common vocabulary, and broad strokes of the language, but it might not be able to replace the knowledge about nuances in language (verbal and nonverbal), changing idiomatic expressions, and other things that an actual human would know.
I think AI learning partners would be helpful for my niece and others her age who struggle with math concepts. An AI learning partner would exhibit unlimited “patience” and thus make learners feel comfortable to ask questions that they might not raise in class in front of their peers. I love the idea of having AI learning partners in elementary school classrooms, especially, and don’t feel that teachers’ roles would be in danger, as many fear. If anything, this technology could be very handy for teachers with oversized classrooms who find it hard to give each student the attention they need.
I would invest in this venture.
Edith
Hi Edith,
I really appreciate the balance of your response. Your language learning example stands out because it nicely highlights one the limitations we often overlook. AI can certainly support vocabulary, grammar, and practice conversations, however culture, humour, context, and nuances of language are much more difficult to capture.
I also like the point you make about AI’s “unlimited patience”. That comes up frequently in the research around AILP’s. Many learners are more willing to ask questions, make mistakes, or revisit concepts when they know they won’t be judged. For students who require additional practice or are reluctant participants, this can be meaningful advantage.
When I think of an AI learning partner, I picture something that can explain ideas, ask helpful questions, and adjust its support when I am stuck. That feels different from a search engine or chatbot because it would work through the learning with me instead of just giving me an answer. When I have learned something difficult in the past, that kind of support might have helped me break it down or see it in a different way. At the same time, I would not want it to remove all the struggle, because working through challenges is often part of learning. By the end of the lesson, I hope to better understand how AI learning partners can support students without making them too dependent on the technology.
Hi Monique,
Love the distinction you make between an AILP and a traditional chatbot/search engine. I think that difference is easy to overlook, but it’s one of those ideas our group hopes participants would get to explore in the OER.
I agree with your point about not wanting AI to remove all of the struggle. Productive struggle is an important part of learning, and one of the biggest design challenges for AILP’s is learning when to provide support and when to encourage learners to persist a little long. Finding that balance is difficult, and that’s one of the factors that will ultimately determine the strength of these tools, or create dependency.
1. When you think of an AI learning partner, what do you imagine it does, and how is that different from a search engine or chatbot? :
When I think of an AI learning partner, I think of something like Khanmigo. I know that these apps are a lot more comprehensive with their functions, and they really guide the participant in getting to the answer but do not give them the answer directly as a chatbot would. They use certain question prompts that follow the Socratic method, and help participants come to a conclusion while also building their own critical thinking along the way. An AI learning partner could also cater to the participants learning styles by presenting information in a layout that works for the user/in a way that the user can understand the information better.
Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable? :
If I am considering the help of AI, it would be that it has helped me understand difficult information that previously might not have been accessible to me, if I did not have AI’s help. For example, sometimes the readings in university (especially the ones I have no background knowledge in) are dense. Tools like Notebook LM have helped synthesize the information for me and really helped me understand the key points of dense articles. However, there are times when I have learned difficult concepts in math for example, by trial and error. I wonder if I had AI’s help to learn those math concepts in the past, would I have built up an understanding of those math concepts just as well? or would have I used the AI as a shortcut? There is a reward that comes when you struggle with something, and overcome it by yourself to finally understand it in the end without anyones help.
By the end of this lesson, what is one question you hope to answer about AI learning partners or their role in education?:
What are all of the AI learning partners? Are there any that I do not know about and will be presented on in this OER?
Hi Gerta,
I appreciate that you connected AILP’s to tools you’ve already used/experienced in your practice (e.g., Khanmigo and NotebookLM). Those examples illustrate how AILP’s differ from traditional chatbots by focusing more on questioning, scaffolding, and helping learners build understanding rather than giving answers.
You also make a strong point on productive struggle. One of the ongoing questions in educational technology is how we provide enough support to reduce unnecessary learner frustration without removing the cognitive effort and leads to deeper learning. As you mentioned, there can be real value in working through a challenging concept independently before everything “clicks”. Those are the “a-ha!” moments in teaching we want our students to experience.
The market for AILP’s is evolving so quickly that new tools seem to be popping up every month. While we couldn’t cover every platform, our team intentionally selected a range of examples that demonstrate different approaches and design philosophies rather than focusing on just one product.
Think about a time when you learned something difficult. How might an AI learning partner have helped or possibly made the experience less valuable?
A few years ago, I was assigned a new position as a Grade 9 Math and Science intervention teacher. Every student in the class had a designated learning disability and/or behavioural designation, and the program itself was a pilot project for our school division. Because no one had taught it before, there was no mentor to guide me through the first year.
Looking back, an AI learning partner would have been an incredible coach. Rather than simply giving me answers, it could have helped me interpret assessment data, suggest evidence-based intervention strategies, challenge my assumptions, connect me with relevant research, and prompt me to reflect on my instructional decisions. Having an AI thinking partner would have made the learning curve much less overwhelming.
At the same time, I would not have wanted AI to replace the struggle. My greatest learning came from experimenting, making mistakes, reflecting on what worked, collaborating with colleagues, and building relationships with my students. Those experiences developed professional judgment in ways AI never could. For me, the most valuable AI learning partner is one that extends my thinking rather than replaces it, helping me become a better educator while ensuring the learning remains my own.
Hi Cherri-Lynn,
I really love how you grounded your response in your own teaching experience. Your example highlights one of the ideas we found repeatedly in our research, that AILP’s are often most valuable when they function as thinking partners rather than answer providers.
I also think you make a great point that your professional judgement developed through experimentation election, and relationships with colleagues and students. That aligns nicely with the principle of extending thinking rather than replacing it, which we explore in the Educational Value section of our OER.
I also wonder whether the role of an AILP changes as expertise develops. Do you think a beginning teacher vs. one with more understanding of learning and pedagogy would benefit from the same kind of AI support, or should an effective AILP adapt its role as a learner’s experience grows? I’d be interested to hear your thoughts!
That’s a great question, Sam, and one I’ve thought about often. I do think a teacher’s experience and pedagogical knowledge influence how effectively they can use an AI learning partner. An experienced teacher brings a depth of professional judgment that AI simply can’t replicate.
One way I think about it is through the concept of chunking. At the end of a lesson, an experienced teacher sees “clean up” as a single task, while a beginning teacher may see 15 or 20 separate tasks that all compete for attention. I think pedagogy works in a similar way. A teacher who understands concepts like teacher clarity, formative assessment, or cognitive load is better able to evaluate AI’s suggestions, recognize what is most important, and decide what will actually improve student learning.
That makes me think an effective AILP shouldn’t provide the same support to every teacher. Instead, it should adapt as a teacher’s expertise grows—offering more guidance and scaffolding to novices while serving more as a reflective-thinking partner for experienced educators. Ultimately, I think it’s that professional judgment, developed through experience, that gives teachers the edge over AI.
I think the idea of an AI learning partner is interesting and potentially helpful for society. When generative AI first came out, I would sometimes think about how it has the potential to freely support students who otherwise wouldn’t have the means to hire their own tutor.
I see AI learning partners as a valuable resource that can help students digest information in a way that truly tailors it for the user.
I do have some underlying ethical, representation/voice, and environmental concerns with these technologies, but I also see the great potential. I suppose my big question would be what the projected long-term effects of personalized AI learning partners will be over time. I wonder how it will change the educator profession and what it means to go to school.
Hi Kane,
Thanks for sharing your perspective and adding to the conversation! I appreciate the point you make about AILP’s potentially increasing access to personalized support for learners who may not otherwise have access to tutoring. I think opportunities like this can broaden educational access, and certainly one of the reasons the AI market is growing so quickly.
I also think it’s important that you connected this potential with ethics, representation, and environmental impact. One thing we hope participants take away from our OER is that designing an AILP isn’t just about making them capable, but making intentional decisions that reflect educational values and priorities.
In terms of the long-term impact on education, do you see AILP’s eventually changing the role of the educator, or do you think the greatest value will be in supporting educators and preserving the importance of human relationships with learning?
By the end of this lesson I would love to have an idea of how to properly implement AI learning partners into the classroom. Throughout the MET program I have had amazing opportunities to learn about the potential of incorporating chatbots and customized AI assistants. The challenge that seems to always come up is the viability of truly incorporating these tools into the classroom. So often opportunities are locked behind paywalls or require a large investment of time and skill to build them into the classroom.
AI is changing the landscape of education and in many instances, I think teachers are way behind students when it comes to how to use these tools. However, while students may know how to access them, they don’t necessarily know how to utilize them effectively without compromising the learning process. I would love to learn more on the viability of using AI learning assistance in my classroom.
I’m really looking forward to working through your project!
Thanks for sharing your thoughts, Megan!
I really connected with your point about the gap between what’s possible with AI and what’s practical to implement in the classroom. While AI Learning Partners offer exciting opportunities, factors like cost, time, technical expertise, and access can all influence whether they’re realistic options for educators.
I also think you make a great point in observing that many of us are comfortable accessing AI tools but still need support in using them effectively. A huge part of my work this year has been partnering with schools and teachers across my division to explore how AI can be used to enhance differentiation, as that continues to be one of our division’s biggest priorities and challenges.
Do you think the biggest barrier to implementing AI Learning Partners is the technology itself, or is it helping educators develop the confidence and pedagogical knowledge to use these tools effectively?
Throughout our OER, our group wanted to explore how thoughtful design decisions can encourage AI to foster thinking, reflection, and learner agency rather than simply becoming an answer-providing machine.
I have to say this OER is incredibly well put together!! The flow of the course, moving from stakeholder pain points to the design sliders and then using research principles to refine our models, really forced me to think about how an AI partner functions and how I can apply this to my day-to-day work.
My favorite part was the design trade-offs simulation activity. Playing around with the different options made it so clear how tiny adjustments to help levels, tone, and memory completely change the user experience. In corporate training, this is an incredibly useful framework. It shows exactly how we can build deeply personalized pathways, like giving a veteran manager and a new hire completely custom tracks, without accidentally stripping away the productive struggle employees need to actually learn.
However, because every section asks you to navigate complex frameworks, read dense predictions, and complete multi-layered tasks, the cognitive load gets pretty heavy.
This training does an amazing job of shifting the conversation away from passive task automation. It puts the focus right back where it belongs: on protecting human agency, fostering intentional dialogue, and designing for authentic learning outcomes.
Q: When you think of an AI learning partner, what do you imagine it does, and how is that different from a search engine or chatbot?
To me, a traditional chatbot or search engine is purely transactional because you type a query and it dumps out an immediate, direct answer. An AI learning partner functions much more like a collaborative coach. Instead of doing the heavy lifting for the learner, it maps out a path using hints and Socratic dialogue to guide them to their own conclusions. It focuses on long term comprehension and skill building rather than just helping someone tick a box or complete a task quickly.
Q: By the end of this lesson, what is one question you hope to answer about AI learning partners or their role in education?
As a corporate designer, my biggest question is how we can realistically roll out these deeply personalized, long term memory architectures without running into major data privacy and corporate compliance bottlenecks. I am really curious to see what the research suggests about balancing hyper individualized tracking with administrative data liability.
Hi Erika,
Thanks so much for taking the time to share your feedback on our OER, Erika! We’re glad to hear the progression resonated with you. As a group, we spent a lot of time trying to ensure our final product moved beyond just discussing AILPs and wanted participants to experience the kinds of design decisions involved in making one.
I especially love the observation you make about the design trade-offs activity. Your example of designing different learning pathways for a veretan vs. new hire illustrates how personalization is about providing the right kind of support of where someone is in their learning journey.
I also appreciate your feedback about the cognitive load. We included multiple opportunities for reflection and application because we wanted participants to critically examine each design decisions rather than move through the activities too quickly. That said, it’s valuable to hear the experience felt for you from a learner’s perspective.
I’m curious though, if you were revising the OER, where do you think simplifying of streamlining would work best without sacrificing depth of learning? Would love to hear your thoughts!
When you think of an AI learning partner, what do you imagine it does, and how is that different from a search engine or chatbot?
Similar to what others have mentioned, I envision an AI learning partner as a tool that supports learning through ongoing interaction, reflection, and personalized feedback rather than simply providing information as a chatbot or search engine would. Specifically, I think AI learning partners have significant potential in scenario-based learning, where the goal is not always to identify a single correct answer but to explore decisions, perspectives, and possible outcomes.
My perspective is likely influenced by my familiarity with scenario-based learning, but I can see strong applications in areas such as career counselling, where an AI partner could simulate interview scenarios and provide tailored feedback based on a learner’s responses. Similarly, in situations involving complex ethical decisions, an AI learning partner could guide students through different perspectives, ask follow-up questions, and encourage deeper reflection on their reasoning.
Thanks for sharing your perspective, Jane! I really appreciated your point about AILPs being particularly valuable in scenario-based learning. I think those kinds of environments are a great example of where AI can move beyond providing information and instead encourage learners to explore different perspectives, justify their decisions, and reflect on their reasoning.
Your examples of career counseling and ethical decision-making also stood out to me because they highlight that many learning situations don’t have a single “correct” answer. Rather, the value comes from dialogue, reflection, and feedback that help learners refine their thinking over time.
I’m curious whether you think an AILP should challenge a learner’s reasoning by presenting an alternative perspective, even if their original response is reasonable. Could that deepen reflection, or do you think it might become frustrating for learners?
Using AI to Skip the Embarrassment of Not Knowing
Recently, I relied heavily on AI to create a video in Canva. Honestly, it saved me hours of trial and error. Throughout the experience, I felt so supported and encouraged that it was easy to forget I was interacting with AI rather than a human coach.
Ironically, one of the biggest barriers I faced was that everyone I work with seemed to be a Canva expert, while I had never really embraced the platform. I was reluctant to ask for help because I didn’t want to appear inexperienced. AI provided a judgment-free space where I could ask as many questions as I needed, make mistakes, and learn at my own pace. It became the learning partner I needed in that moment.
At the same time, the experience highlighted an important trade-off. As Brené Brown reminds us, there is power in vulnerability. By turning to AI instead of my colleagues, I gained confidence and new skills, but I also missed an opportunity to build relationships through shared learning. The experience reminded me that while AI can be an exceptional learning partner, it cannot replace the growth that comes from being vulnerable with other people.
I’m curious whether others have experienced something similar. Has AI ever given you the confidence to learn something you were hesitant to ask another person about, or do you think it risks replacing the very human interactions that help us grow?
I can definitely relate to your experience. One things I appreciate most about Ai is having a space to ask questions and work through ideas without having to necessarily worry about looking inexperienced.
However, at the same I think the point you make about vulnerability is an important one. Some of the best professional learning I’ve had or given has come from the conversations with colleagues, and I wouldn’t want Ai to replace those opportunities for collaboration and relationship-building. For me, the greatest value of an AILP is when we can keep human-centered approaches at the forefront.
To answer the first question, I think of a learning partner as an aid in learning something new. In my understanding, it is not something you go to get a default answer that is generalized or not explained, but helps to walk one through a process of prior knowledge assessment and application. It is meant to aid in the learning process, not to give a point-blank answer like a Google search or Snapchat AI.
HI Emily,
Thanks for sharing your thoughts. I love how you describe an AILP as something that guides learners through a process rather than just providing answers. That distinction between supporting learning and delivering information is one the key ideas that helped shape our OER. I’m curious to hear your thoughts and feedback after exploring the activities – did they change how you think of AILPs?
When I’ve learned something difficult, an AI learning partner would have helped by breaking concepts into structured, bite-sized pieces and guiding my thinking through probing questions, the Socratic method others have mentioned. Better yet, if it also reinforced learning by testing my knowledge from multiple angles and perspectives, it could deepen genuine understanding and deliver more value.
But I think the delivery mode also matters. If responses are overly dense, lengthy or otherwise inaccessible, I risk accepting them at face value rather than genuinely engaging, which can build a dependency that looks like performance but bypasses learning.
I really appreciate the point you make about AI breaking complex ideas into manageable pieces while still encouraging learners through probing questions. That balance between support and preserving cognitive work of learning is one of our central design ideas and something we explore in detail in the trade-offs section.
I also think your point about deliverance is especially information. If AI overwhelmed learners with lengthy or inaccessible responses, it can actual become a barrier for learning, rather than a support. Do you think giving learners more control over how AI responds would address this challenge? For example, choosing the level fo detail or guidance an AI tool could give.