From Digital to Spatial: How AI Is Changing the Economics of Immersive Learning

Spatial learning has never suffered from a shortage of compelling use cases. Anatomy, chemistry, engineering and many other subjects are inherently three-dimensional (3D), yet students are still often asked to learn and understand them through text, two-dimensional diagrams or flat video. The main obstacle has been economics: creating enough high-quality interactive 3D content to cover an ordinary curriculum that has traditionally required specialized artists, developers and considerable production time. Artificial intelligence (AI) is beginning to change this with new forms of spatial computing making immersive experiences easier to deliver across a wider range of devices and learning environments.

The opportunity is not to replace textbooks, lectures, physical laboratories or teachers. It is to match the physical representation to the concept. A student can read about the structure of a molecule, examine an illustration of the human heart or study a diagram of an engine, but in each case, the learner must mentally reconstruct relationships between objects, structures and processes. Spatial learning can reduce some of that cognitive translation by allowing students to inspect those relationships directly and interact with them from different, life-like perspectives.

From Digital Information to Spatial Understanding

The move toward spatial learning is part of a broader evolution in educational technology. Digitizing a textbook made information easier to access but did not fundamentally change how that information was represented. Interactive software added participation, and simulations added experimentation. Spatial experiences can go one step further by changing the representation itself, particularly when the subject is inherently spatial.

Consider anatomy. Recognizing the structures of the human heart is only the beginning; students must also understand how those structures relate to one another, where they sit within the body and how they function as part of a larger system. The same challenge appears in chemistry, where molecular geometry matters, and in engineering, where understanding the interaction between components can be as important as identifying the components themselves.

A growing body of research suggests that this is more than a theoretical advantage. A 2024 meta-analysis published in Anatomical Sciences Education examined 27 experimental studies involving 2,199 health-sciences students and found that extended reality (XR) technologies produced greater knowledge gains than traditional approaches. When XR was used as a supplemental resource alongside conventional instruction, this effect increased. Knowledge performance using XR also outperformed textbooks and atlases, with 80% of students who used XR reporting that they considered the technology useful for learning anatomy.

A separate meta-analysis of randomized trials reached a similar conclusion, while also finding substantial variation across studies. That caveat matters: spatial technology is not automatically educational simply because it is immersive. Its value depends on instructional design, the quality of the representation and how well the experience is integrated with the rest of the lesson.

The same principle extends across STEM. Physics involves forces and motion that can be difficult to observe directly. Earth and space sciences deal with phenomena that may be too large, too small, too distant or too dangerous to reproduce physically. In each case, a good spatial representation can make relationships that would otherwise remain abstract visible.

Spatial environments can also make experimentation more accessible. Instead of simply being shown the result of a process, students can manipulate variables, test hypotheses and observe the result. A simulation lets a learner make a mistake, change an assumption and try again without consuming physical materials or introducing the same safety and logistical constraints as some laboratory activities.

Research into complex biological systems illustrates the potential particularly well. In a study of 3D, haptic-enabled virtual reality for learning about the human heart, examined whether an interactive 3D experience could support secondary students learning about cardiac anatomy and physiology.  The researchers concluded that 3D, haptic and VR technologies can provide robust representations and student-driven interactions with complex biological systems, while also noting that younger students may require additional support to connect new information to broader systems. 

These examples also show why immersive education should be thought of as a broader category than VR alone. The experience may take place in a headset, through augmented reality, on a screen-based glasses-free 3D display or through another interface. The relevant question is not which form factor is universally best, but which one helps students understand a concept that would otherwise be difficult to visualize or experience.

As that category expands, the most important shift may be from isolated demonstrations toward experiences that fit naturally into ordinary teaching. That requires not only compelling hardware, but also enough content, curriculum alignment and ease of use that spatial learning can become another everyday instructional tool.

AI Is Changing the Economics of Spatial Learning

If spatial learning has shown promise for years, why has it remained relatively specialized? One of the biggest obstacles has been content with economics. A high-quality interactive lesson can require subject-matter expertise, instructional design, 3D modeling, software development and significant production time. That model works for selected flagship experiences; it is much harder to scale across an entire curriculum.

Artificial intelligence could begin to change that equation. Generative AI is already being adopted by educators for practical tasks such as lesson planning, research and content development. According to Gallup, six in 10 teachers report using AI for their work. Among lower-secondary teachers, OECD’s 2026 Digital Education Outlooks found that 57% agreed that AI helps them write or improve lesson plans, while 72% believed AI could harm academic integrity by enabling students to pass off work generated by others as their own.

Those numbers illustrate both the momentum and the challenge surrounding AI in education. Teachers are already experimenting with the technology, but the most useful applications are likely to augment rather than replace professional judgment. The same principle should guide the use of AI for spatial learning: automate friction, not pedagogy.

AI is beginning to automate parts of the spatial-content pipeline that once required specialized workflows. Depending on the material, models can assist with depth estimation from 2D imagery, segmentation, 2D-to-3D conversion, generation of 3D assets from text or images, reconstruction of scenes from photographs and video and adaptation of visual material for different spatial interfaces. None of these techniques are perfect, but together they change the starting point. A textbook illustration, scientific diagram, photography or video can increasingly become raw material for an interactive spatial experience rather than requiring a team to build every asset from scratch.

That shift matters because scaling education is largely a problem of reuse. If teachers and instructional designers can start from materials they already trust, while AI handles more of the repetitive technical work required to make those materials spatial, immersive learning becomes easier to imagine at curriculum scale rather than as a collection of expensive one-off demonstrations.

AI may also make the same spatial lesson more adaptable. A difficult molecular structure, for example, could be presented first in a simplified form and then progressively revealed in greater detail. Another student might spend more time manipulating the same model or receiving additional explanatory context. The educator still defines the learning objective; AI can help vary the path used to reach it.

That possibility needs to be carefully planned and approached. UNESCO’s guidance on generative AI in education and research calls for a human-centered strategy that protects human agency, addresses privacy and safety concerns and ensures that AI systems are evaluated for their ethical and pedagogical suitability. UNESCO also emphasizes that AI should support human intelligence rather than displace it. 

The objective, then, should not be to automate education. It should be to lower the cost and expertise required to create useful spatial experiences while keeping educators responsible for what is taught, how it is taught and how learning is assessed.

Bringing Spatial Learning Into the Everyday Classroom

Content creation is only one part of the equation. The other is how students access those experiences.

For much of the history of immersive technology, the most visible path to spatial experiences has been the VR headset. Headsets can create powerful levels of immersion, but classroom deployment also raises practical questions around cost, maintenance, accessibility, device management and how easily a teacher can move between individual and group instruction.

This is not an argument against head-mounted devices. For some lessons, full immersion may be exactly the right tool. But spatial learning will be more useful if educators can choose among multiple delivery modes. Screen-based, headset-free 3D experiences, for example, can preserve many of the benefits of spatial visualization while reducing friction for teacher-led or collaborative use. The question is not headset versus display; it is matching the interface to the teaching method.

That distinction matters because classroom learning is inherently social. Students need to ask questions, compare observations and look back to the teacher or to one another. Spatial technology is most effective when it supports that interaction rather than becoming the destination itself.

One useful real-world example is zSpace, whose education platform combines screen-based, headset-free 3D hardware with curriculum-aligned software, simulations and professional learning. The important lesson is less about any one product than about the delivery model: spatial learning is easier to integrate when the technology behaves like part of the classroom rather than requiring every student to enter a separate virtual environment.

The same principle extends into professional and medical learning. At UHealth – University of Miami Health System, physicians are using Eonis Vision, developed by Avatar Medical with Barco, to turn CT and MRI scans into interactive, patient-specific 3D anatomy during consultations. The current deployment is focused on patient understanding and clinical communication rather than formal medical education, but it illustrates the same educational principle: complex anatomy can become easier to discuss when two people can examine the same spatial representation together. The same approach has clear potential for students and trainees learning anatomy from real clinical data.

Research also suggests that immersive and spatial technologies are often strong as complements to conventional instruction rather than replacements for it. The anatomy evidence cited earlier found particularly strong gains when XR was used as a supplemental resource. That supports a model in which spatial experiences are woven into a broader lesson alongside discussion, reading, physical experimentation and teacher guidance.

The Next Generation of Digital Learning

The first wave of digital education made information easier to access. The next wave will make information spatial. That is a more fundamental change than simply putting another type of content on a screen, because it changes how students can interact with the information itself.

Instead of only reading about a complex structure, students can examine it. Instead of looking at a diagram of a system, they can manipulate it. Instead of memorizing the steps of a process, they can change variables and observe the consequences. These experiences can help bridge the gap between an abstract explanation and a learner’s internal model of how something works.

The traditional classroom does not disappear in the future. Teachers remain essential, and foundational skills such as reading, writing, discussion and critical thinking remain just as important. The opportunity is to match the learning format to the nature of the concept: some ideas are best taught through lecture or text, some through physical experimentation and some become much easier to understand when students can see how the components relate in space.

AI and spatial computing can make that last category much more accessible. AI can reduce the time, cost and specialized expertise required to transform existing materials into interactive experiences, while advances in spatial hardware can make those experiences available across more classroom settings and devices.

The classroom itself may therefore look surprisingly familiar. Teachers will still teach and students will still collaborate. What changes is the nature of some of the information inside that environment: students can increasingly explore, manipulate and understand it from multiple perspectives rather than only consume a fixed representation.

That is the larger promise of the shift from digital to spatial learning. It is not about replacing the ways students learn today but expanding the number of ways they can understand what they are being taught. If AI can continue lowering the barrier to creating spatial content, immersive learning has a clear path from specialized experiences to an ordinary part of the educator’s toolkit.