The first generation of Chatbots and concierge-style virtual assistants like Siri and Alexa have evolved into a competitive arena of client-focused virtual assistant (VA) companies, including Wishup, Athena, Belay and Boldly, where the VAs offer private, personalized information and support, and can also perform a wide array of tasks, either autonomously or under direction.
Next up are Learning Partners (LPs), such as AskBecky, that offer academic coaching and support, at this point mostly in hybrid form (human coach plus AI). However, it is widely believed that 2026 will introduce the first fully virtual Learning Partners, building on strong foundations demonstrated by NotebookLM, Khanmigo and others, to offer students and teachers their own personal – ostensibly lifelong – virtual Learning Partners.
Opportunity Statement:
Learning Partners are inevitable, and prospectively powerful, and will need lots of oversight as they develop. How will they best fit in to every learner’s formal, informal and social learning environments?
Sources:
DeepSeek, Gemini
The best teachers are the ones who understands a learner’s specific strengths and weaknesses, motivations, interests, goals, learning preferences, and inspirations. Learning Partners, then, seem an inevitable next step in education, as they would have the capability to understand and utilize the learner’s personal idiosyncrasies to tailor learning in ways that perfectly suit the student.
This technology therefore has the potential to drastically improve the depth of learning in future students. It could present all material in a way that the learner would find engaging, track where the learner most needs to improve, and tailor coaching sessions to best help this particular learner with this particular lesson on this particular day. Deep learning happens when a person is actively engaged and interested, and I feel this technology could be a game-changer in presenting effective, adaptive learning opportunities that appeal to an audience of one.
What I find most interesting about Learning Partners is that they move AI beyond simply being a “tool” towards becoming an ongoing cognitive companion integrated into the learning process. I agree that systems capable of adapting to learner needs, interests, and challenges could create far more personalized and responsive learning experiences than what many students currently receive.
However, I also think it’s important to consider agency and dependency on this in learning. If students (and ourselves) increasingly rely on AI systems to summarize information, generate ideas, guide problem-solving, or structure thinking, how does this shape critical thinking, creativity, and knowledge construction over time?
Educators to me play an essential role in helping students navigate these technologies rather than outsourcing cognition to them. But, I also wonder whether schools are currently equipped to navigate this shift effectively? Much of the professional development around AI still focuses primarily on surface-level tool use and efficiency, rather than deeper conversations around cognition, ethics, critical digital literacy, and how these technologies reshape learning.
For me, Learning Partners feels less like a distant future thing, and more of an emerging reality that schools are just beginning to navigate.
Hi Sam,
I think you pose a really valuable question in your second paragraph, but I see it less as a skeptical critique of the technology and more as an opportunity to strengthen the factors you mentioned (creativity, critical thinking, etc). If the team designing the Learning Partner is directed by knowledgeable educators (as opposed to businesspeople), it would hopefully be coded to develop those skills intrinsically within its procedures, for example by intentionally leading learners to deeper understanding rather than simply spewing facts and generating assessments.
I found the idea of Learning Partners especially interesting because it suggests a shift from AI being a tool that learners occasionally use towards becoming an ongoing companion within the learning process. While tools like NotebookLM and Khanmigo already show early forms of this, imagining lifelong personalised learning support raises bigger questions about what learning could become.
At the same time, I wonder whether the greatest challenge is not technical capability but maintaining learner agency. If Learning Partners become increasingly capable of summarising information, suggesting directions, and guiding thinking, there is a risk that learners begin outsourcing parts of cognition rather than strengthening them.
From my experience working in school eLearning and teaching contexts, I see strong potential for Learning Partners to support differentiation and provide more timely feedback, especially in large classes where personalised support is difficult (very common in the Macao context). However, I think educators will become even more important as facilitators who help learners question, reflect, and remain intentional about how they engage with AI.
Perhaps the future question is not whether students will learn with Learning Partners, but how we design these relationships so that support does not become dependence.
AI-powered learning partners are already here and their wide-spread adoption is inevitable. What is still unclear is whether the makers of these products will be incentivized to build them in a way that works to perserve learners’ critical thinking skills or in a way that erodes those skills in favour of dependance. Either way, their impact will be significant. I would argue that content on demand is a subset of learning partners.
The continuity problem is what makes Learning Partners (LPs) most professionally urgent for me. As a personalized private math instructor working in K-12 settings, the primary challenge I contend with is the continuity gap between my curriculum-aligned sessions and classroom instruction, where students are without the adaptive, individualized support I provide. An LP that fills that gap, responsive to where a student actually is and aligned with the approach established in our sessions, could be transformative. Fully virtual LPs, expected to emerge imminently, also seem worth examining for how they may operate alongside human educators and other LPs implemented in hybrid configurations, in and out of the classroom.
Building on a point others have raised, the LP design principle that matters most to me is educator-orchestration, where human tutors and teachers set the instructional frame, and the LP operates within it rather than substituting for the relational judgment and contextual empathy a human brings. I think the version worth building is an adaptive presence that knows the student’s current learning edge, scaffolds precisely and in real time, and defers to the human educators appropriately, rather than simply a chatbot that fills academic gaps on demand.
I am fascinated by how AI technologies are moving past being simple tools and are developing into learning partner technologies catered toward individual users and use cases. I believe learning partners will become fundamentally important in the future because they represent a major departure from static educational tools toward dynamic systems that have the ability to adapt to individual users’ needs across various formal, informal, and social learning contexts. With recent technological advances like Retrieval-Augmented Generation and platforms like NotebookLM making this tech more accessible to the average user, I am curious how learning partner technologies will reshape learning environments and learners’ self-learning capacities. Ultimately, as this technology continues to be implemented, I feel its true value will be realized as educators learn to strategically leverage their own instructional methodologies, lessons, and resources alongside these learning partners to foster meaningful learning collaborations and contexts.