The Wizard of Oz opened at the Sphere this weekend, and I was able to get glimpses from YouTube and other media coverage. The $100 million AI-enhanced spectacle is genuinely breathtaking. Dorothy's tornado whirls around you in 16K resolution while wind effects make you grip your seat, and the Yellow Brick Road stretches infinitely across the venue's 160,000-square-foot wraparound screen.
While the headlines focus on Google DeepMind's AI wizardry, I think the real magic comes from the 2,000 human artists, engineers, and researchers who spent two years making impossible creative decisions.
Google's technical achievement is undeniable. Their AI models (Veo 2, Imagen 3, and Gemini) performed what Dr. Steven Hickson calls "wildly innovative" computational feats. They upscaled grainy 1939 footage to ultra-high definition, used "outpainting" to extend scenes beyond their original frames, and generated new character performances to fill the Sphere's massive canvas.
I found the outpainting technique particularly impressive. When the original film shows Dorothy facing the tornado, AI expands the frame to reveal more of the Kansas landscape, Dorothy's legs, and swirling debris that Victor Fleming never filmed. When Uncle Henry walks into frame halfway through a scene, you now see him listening by the front door before he decides to cross the room.
The technical challenge was staggering. "There are scenes where the scarecrow's nose is like 10 pixels," Hickson noted. "That's a big technological challenge."
But the technology tells only part of the story.
What the media coverage misses is the extraordinary human curation that made this AI enhancement possible. The team didn't just feed the original film into an algorithm and hope for the best. They assembled what amounts to a comprehensive archive: shooting scripts, production illustrations, set photographs, architectural plans, costume designs, and scores of reference materials from the MGM archives.
Through a process called "fine-tuning," these archival materials trained the AI models on the specific visual language of Fleming's Oz. The system learned not just what Dorothy looked like, but how she moved, how the Kansas farmhouse was constructed, and even the focal lengths used for specific scenes.
This curatorial work required deep film expertise. When expanding a scene of the Cowardly Lion pouncing on his companions, human artists had to make strategic decisions: Which architectural details from the original set design should inform the expanded background? How would Fleming have staged the wider shot? What emotional tone should the new framing support?
The project raises fascinating questions about creative authenticity. Warner Bros. and the production team emphasize their commitment to Fleming's vision, studying his shot lists and meticulously plotting the world beyond the original 4:3 frame. Every AI-generated addition underwent review to ensure continuity with the 1939 spirit.
Yet critics rightly point out that we're seeing Uncle Henry in positions he never occupied, Dorothy in landscapes Fleming never imagined, and character interactions generated by algorithms trained on contemporary data. The project has sparked intense debate among film purists, with some calling it "artistic butchery" and "the death of cinema"—criticism that highlights broader concerns about AI's role in reimagining classic works.
This tension reveals something important about AI's role in creative work. The technology excels at extrapolation, extending existing patterns with remarkable fidelity. But the creative choices about which patterns to extend, how to maintain emotional authenticity, and when to preserve original limitations required distinctly human judgment.
What makes this project educational isn't its AI brilliance, but its demonstration of human-AI collaboration at scale. The most successful moments aren't where AI replaced human creativity, but where it amplified human decision-making.
Producer Jane Rosenthal, co-founder of the Tribeca Film Festival, brought expertise that no algorithm could replicate: understanding which scenes would benefit from expansion, which performances should remain intimate, and how to balance spectacular visuals with narrative coherence. Google's Buzz Hays leveraged his decades of entertainment industry experience to guide technical choices toward dramatic impact.
The Sphere team made strategic choices about when to intervene. Rather than automating the entire enhancement process, they maintained human oversight at every creative decision point. When AI generated new character movements, human artists evaluated whether they felt authentic to the original performances. When algorithms extended backgrounds, production designers ensured they supported Fleming's visual storytelling.
This project offers a compelling model for educators integrating AI into their classrooms. The Sphere team didn't ask "How can AI create The Wizard of Oz?" They asked "How can AI help us realize creative possibilities that weren't technically feasible in 1939?"
That reframing is crucial for educational technology. Instead of asking "How can AI grade essays?" or "How can AI teach students?" we should ask "How can AI help teachers provide feedback that wasn't practically possible before?" or "How can AI help students explore learning pathways that traditional resources couldn't support?"
For Teachers Using AI Tools: Students could use AI to explore alternative story structures while maintaining responsibility for thematic coherence and character development. Teachers could use AI image generators to create additional context around primary sources (showing what a WWI battlefield might have looked like beyond the edges of a historical photograph) while ensuring students distinguish between documented evidence and AI-generated interpretation. Math teachers could use AI to generate unlimited practice problems at specific difficulty levels, then focus their expertise on helping students understand problem-solving strategies and mathematical reasoning. Science teachers could use AI to create visual models of molecular structures or astronomical phenomena that are impossible to observe directly, while maintaining focus on the underlying scientific principles and evidence.
The Curatorial Imperative: The key insight is curatorial: the quality of AI enhancement depends entirely on the quality of human educational direction. The Sphere team's success came from their deep knowledge of Fleming's work, their understanding of cinematic storytelling, and their commitment to preserving what made the original magical. Similarly, successful educational AI implementation requires teachers who deeply understand their subject matter, pedagogical best practices, and what makes learning transformative for their specific students.
The Human Intelligence Multiplier: Standing in the Sphere, watching Dorothy click her heels in impossible visual clarity, you realize this isn't really about AI at all. It's about what becomes possible when technology serves human creative vision rather than replacing it.
Google's algorithms may have generated the pixels, but 2,000 people made 10,000 creative decisions about which pixels mattered. They chose which archival materials to prioritize, which emotional moments to expand, and which original limitations to respect. They balanced technical possibility with narrative necessity, spectacular visuals with intimate storytelling.
For educators, this principle is foundational: AI tools are amplifiers of human expertise. The technology succeeds when teachers use their deep understanding of learning, their knowledge of individual students, and their pedagogical judgment to direct AI toward specific educational outcomes.
The Sphere's Wizard of Oz succeeds not because AI can perfectly upscale vintage film, but because talented people made thoughtful choices about when and how to deploy that capability in service of Fleming's original vision.
The most effective educational AI implementations will follow the same pattern: human expertise directing technological capability toward learning goals that couldn't be achieved through analog means alone.
Speaking of following the yellow brick road to better education, Herbie Raad and I explored similar themes about wisdom, courage, and authentic teaching in our 2021 book The Teachers of Oz: Leading with Wisdom, Heart, Courage, and Spirit. The Sphere's AI enhancement demonstrates impressive technical capability, but as we argued in our book, the real transformation in education happens when teachers bring their wisdom, courage, and authentic leadership to their practice—whether they're working with cutting-edge technology or traditional tools.
Never Stop Asking,
Nathan