The AI XR Podcast

Jonathan Rodriguez Cefalu built the hardware that Snap shipped on people's faces — first the camera-only Gen 1 Spectacles, then the Gen 4 display version. His path through Stanford CS, an honors thesis on varifocal display optics, and a startup called Vergence (named after the vergence-accommodation conflict in AR) led him to Snap, and then to the problem he is working on now. Preamble AI exists to prevent the worst possible AI outcomes — starting with a class of attack that Preamble was the first to publicly demonstrate: prompt injection.Ted Schilowitz hosted this episode solo. Together, he and Jonathan worked through the architecture problem sitting under every AI assistant being deployed at scale right now: large language models see one token stream. There is no separation between what the developer intended and what an untrusted email or web page is quietly instructing the model to do. With Gemini Spark about to give AI agents access to tens of thousands of emails per user, this is not a theoretical concern. Jonathan's team has a proposed fix — and they have already shaped federal law.The episode also covered the week's XR and AI news: Google I/O announcements, Snap Spectacles Gen 6 details ahead of AWE, Matthew Ball joining Xbox, Anduril's battlefield AR wearable, and AI-generated feature films reaching Tribeca.AI XR News You Should Know:Google unveiled Gemini Spark at I/O — a persistent AI agent integrated across Gmail, Docs, Chrome, and workspace tools, now in beta for paid subscribers. Ted and Jonathan tested pre-Spark Gemini Gmail and found it searched roughly 30 emails when asked to search tens of thousands. "It just got lazy." Both came away cautiously pessimistic about agentic reliability at scale. XREAL Project Aura was also announced — birdbath optics connected via USB-C — solid engineering but not new ground. Android XR is spending heavily for incremental progress.Snap Spectacles Gen 6 is expected to preview at AWE in mid-June at around $2,500. Jonathan led Gen 1 and Gen 4 from the inside and broke down what Snap has genuinely solved: low-energy on-device 6DoF tracking, hand tracking, spatial mapping, and multi-device sync. The Lens Studio developer ecosystem is healthy, with a Unity scene auto-converter recently open-sourced. His read: Snap does more with less. Meta does less with more, and it traps talented researchers — like Douglas Lanman — inside labs where work never ships.Matthew Ball was named Xbox Chief Strategy Officer. Anduril revealed EagleEyes, a battlefield AR wearable with an 84-degree field of view and thermal imaging built for helmet integration. Ted's reaction: scary but fascinating. Both hope the smart people behind it are pointed toward good outcomes.At the AI on the Lot conference in West LA: Amazon previewed an AI-assisted animated children's series called Project Nara. Higgsfield screened "Hell Grind" — a 90-minute AI action film made by 15 filmmakers, 16,000 video generations, and $500,000 total (roughly $400,000 in compute). Paul Schrader came out as pro-AI. "Dreams of Violet," a 75-minute AI feature about Iranian resistance, premieres at Tribeca on June 10 — total cost: $2,000, production time: two months. Jonathan's take: it is sad when AI displaces human creative talent on screen, but thrilling when imagined through AR glasses making a morning commute feel like driving alongside the ocean.Key Moments:[00:00] Ted opens solo — Charlie Fink and Rony Abovitz are out for the summer solstice[02:30] Google I/O: Gemini Spark and what "persistent AI agent" actually means in practice[08:15] Jonathan's Gmail test: asked to search tens of thousands of emails, it searched 30 and quit[14:40] XREAL Project Aura and the state of Android XR — a lot of spend for incremental steps[21:00] Snap Spectacles Gen 6 details: what Jonathan knows from building Gen 1 and Gen 4 from the inside[31:20] Snap vs. Meta: research that ships in the product vs. research that ships in a paper[38:45] Matthew Ball joins Xbox, Anduril EagleEyes, and battlefield AR wearables[44:10] AI on the Lot: Project Nara, Hell Grind, Dreams of Violet, Paul Schrader goes pro-AI[52:30] Jonathan introduces Preamble AI and the mission to prevent worst-case AI outcomes[58:00] The first public demonstration of prompt injection — what happened and why it matters[01:06:15] Why Gemini Spark will be especially vulnerable to prompt injection attacks[01:14:00] Preamble's proposed fix: a reserved token language that untrusted data cannot speak[01:21:30] NDAA Section 1638: the first US law making it illegal to give AI autonomous nuclear control[01:28:45] WarGames, "the only winning move is not to play," and what that means in 2026So what: Every AI assistant being deployed right now — including the ones about to read your email — operates on a flat token stream with no built-in trust hierarchy. Prompt injection is not a future risk; Preamble demonstrated it years ago. The question is whether the companies shipping agentic AI at scale will build protections before the attacks become routine. Jonathan's work is happening on two fronts: the technical architecture and the policy layer. The NDAA provision is proof that the second front can move faster than most people expect.Brought to you by Zappar and Mattercraft. Mattercraft makes spatial web experiences that run in the browser — no app required. Visit mattercraft.io to learn more and start building.Watch the full video on YouTube - https://youtu.be/txUicSGoe98See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Edward Saatchi has been building at the frontier of AI storytelling for a decade—from Oculus Story Studios to Fable (where his AI character Lucy made her own films at Sundance) to his current venture, Amazon-backed Showrunner. His thesis is provocative: AI-generated content is stuck in a four-year rut of short-form experiments with no commercial marketplace, no monetization path, and no artistic value. Creators are working solo, making 10-second clips that can't compete with Rick and Morty or Netflix originals. The solution? Band together, make features and TV shows, and build platforms where creators get paid every time someone remixes their work.Edward's most audacious project proves the point: reconstructing Orson Welles' lost masterpiece, The Magnificent Ambersons (44 minutes destroyed by studio cuts in 1942), using motion-capture actors and AI to seamlessly restore what was erased. The irony is intentional—it's a film about technology destroying beauty, restored by technology. Edward's approach isn't text-to-video slop. It's human performance driving AI synthesis: hire stage actors, capture their performances, use the original cutting continuity as a blueprint, and let AI fill the gaps. The result is cinema-quality work that would cost $100 million traditionally but costs $10 million with AI assistance.In AI XR News This Week: Amazon announces 16,000 layoffs (mostly middle management) while ramping robotics—replacing humans with machines in warehouses. Amazon Go and Amazon Fresh stores close after years of investment; the self-checkout convenience experiment dies. Snap spins off Spectacles AR glasses into a separate business, signaling lack of cash or confidence. Apple and OpenAI both developing AI wearables to launch in 2027, powered by Gemini and Google AI. Google launches Project Genie, a generative AI model that creates fully interactive 3D game worlds you can navigate and remix in real time. Walkabout Mini Golf (one of the 10 most popular Quest apps) lays off half its staff. Atlas V, the acclaimed French VR studio behind Spheres and Battle Scar, pivots to location-based entertainment. Darren Aronofsky launches an AI animated series on YouTube called On This Day.Key Moments Timestamps:[00:05:00] Amazon's 16,000 layoffs paired with robotics expansion; the canary in the coal mine for white-collar work[00:06:00] Amazon Go/Fresh failure: humans reject automated futures when given the choice[00:07:14] Snap spinning off Spectacles; Ted's thesis on AR glasses remaining "exotic," not mainstream[00:10:00] Apple wearables running Gemini + Google AI; the winning formula for wearable AI domination[00:12:48] Walkabout Mini Golf layoffs and Atlas V's pivot; VR right-sizing continues[00:15:25] Google Genie: generative 3D worlds, playable and remixable in real time; Epic should be scared[00:19:11] Edward Saatchi joins: the state of AI video and why there's no marketplace after 4 years[00:22:00] Edward's concern: AI content is "derivative but worse" with no commercial value[00:28:00] The marketplace problem: no buyers, no revenue, no sustainability for creators[00:34:00] Ted's thesis: AI is quietly disrupting VFX and screenwriting behind the scenes[00:44:00] Critters: the proof-of-concept for AI-assisted theatrical animation ($10M vs. $100M traditionally)[00:49:00] Showrunner's business model: creators earn money every time someone remixes their show[00:52:00] The Magnificent Ambersons project: restoring Orson Welles' lost masterpiece with AIEdward makes a case that reads like a manifesto: AI's killer app isn't making derivative work faster or cheaper. It's remix, interactivity, and personalization at scale—letting audiences co-create with AI while creators get paid. His challenge to the industry: hold yourself to "derivative but better" (can you make a better Simpsons episode than the last 15 seasons?) or "original and good" (something from a non-human intelligence's perspective). Until creators band together to make features and TV shows with commercial value, AI video will remain stuck in the trough of disillusionment.This episode is brought to you by Zappar, creators of Mattercraft—the leading visual development environment for building immersive 3D web experiences for mobile headsets and desktop. Mattercraft combines the power of a game engine with the flexibility of the web, and now features an AI assistant that helps you design, code, and debug in real time, right in your browser. Build smarter at mattercraft.io.Listen to the full episode and subscribe to the AI XR Podcast for weekly conversations at the intersection of AI, entertainment, and the future of interactive media. Watch on YouTube.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

What happens when someone who grew up in the Lucasfilm Games golden era decides that today’s AI tools are failing creatives? Mike Levine has spent more than 30 years building at the intersection of games, XR, VFX, and interactive storytelling—and his verdict is clear: the current AI stack is a fragmented, overcomplicated mess that turns directors into prompt engineers.Mike started as a tester at Lucasfilm Games (later LucasArts), working his way into the art department on titles like Sam & Max and The Dig before helping ship live-action Star Wars games such as Rebel Assault and Jedi Knight II. He later built rotoscoping tools used across the VFX industry, collaborated with ILM and Pixar, experimented with mobile AR games for Hasbro and HoloLens, and dipped into crypto gaming—before finally co-founding MovieFlow (now FilmSpark), an AI-native production platform designed so that filmmakers, agencies, and showrunners can move from script to screen without needing a computer science degree.The AI XR news you should know: Apple taps Google Gemini to power Siri, acknowledging that building world-class LLMs in-house makes little financial sense. Meta cuts 10% of Reality Labs, right-sizing its VR bets while pivoting toward wearables. Xreal raises another $100M amid questions about Chinese state influence and data flows. Higgs Field lands $80M at a $1.3B valuation for AI cinematography tools that many filmmakers still find unreliable. Wikipedia signs licensing deals with major AI companies after years of being scraped for free. OpenAI invests $252M in Sam Altman–backed Merge Labs, raising fresh conflict-of-interest questions.Key Moments Timestamps:[00:23:02] From Boston journalist-to-be to accidental hire at Lucasfilm Games[00:26:24] The “test pit” culture at Lucas and how Nintendo experience got Mike in the door[00:28:45] Moving into the art department, learning Photoshop from early legends, and shipping Sam & Max[00:31:15] Live-action Star Wars games: Rebel Assault, Jedi Knight II, and convincing George Lucas[00:34:38] Visiting Pixar with new VFX tools and recognizing the same creative “magic” as LucasArts[00:36:24] Doug Trumbull’s influence on Mike’s sense of cinematic possibility and immersion[00:43:27] The urinal meeting at Magic Leap and what early spatial computing got right (and wrong)[00:49:00] Why most AI tools are “dark ages” for filmmakers: node graphs, 10+ subscriptions, no story view[00:51:00] Building MovieFlow/FilmSpark: story-first, timeline-based AI production for long-form and vertical shows[00:53:00] The Neighborhood Podcast: a 90-second vertical murder mystery as proof-of-concept for AI-native seriesWhen humans can generate shots, scenes, and even entire episodes in minutes, the bottleneck shifts from production to vision. Mike argues that the winning AI tools will be the ones that let directors see their whole story, maintain continuity, and iterate fast—without ever feeling like they left the edit bay for a dev console. His vertical drama collaboration with Charlie, The Neighborhood Podcast, is an early look at what happens when narrative craft meets AI-native pipelines instead of fighting them.This episode is brought to you by Zapar creators of Mattercraft—the leading visual development environment for building immersive 3D web experiences. Build smarter at mattercraft.io.Watch the full episode on YouTube and subscribe to the AI XR Podcast for weekly conversations with the people building the future of AI, XR, and interactive media.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Gary Shapiro has spent decades at the center of the global consumer technology industry, leading the Consumer Technology Association (CTA) and building CES into one of the most important stages for innovation, policy, and deal-making on the planet. In this first episode of 2026, Gary joins Charlie, Rony, and Ted to preview CES, unpack the explosion of AI across every category, and deliver unusually blunt takes on tariffs, China, manufacturing, and U.S. innovation policy. He explains how CES has evolved from a TV-and-gadgets show into a global platform where boards meet, standards are set, and policymakers, chip designers, robotics firms, and health-tech startups all collide.In the News: Before Gary joins, the hosts break down Nvidia’s $20 billion “not-a-deal” with Singapore’s Groq, the stake in Intel, and what that combo might signal about the edge of the GPU bubble and the shift toward inference compute, x86, and U.S. industrial policy. They also dig into Netflix’s acquisition of Ready Player Me and what it suggests about a Netflix metaverse and location-based entertainment strategy, plus Starlink’s rapid growth and an onslaught of “AI everything” products ahead of CES.Gary walks through new features at this year’s show: CES Foundry at the Fontainebleau for AI and quantum, expanded tracks on manufacturing, wearables, women’s health, and accessibility, plus an AI-powered show app already fielding thousands of questions (top query: where to pick up badges). He also talks candidly about his biggest concern—that fragmented state-level AI regulation (1,200+ state bills in 2025) will crush startups while big players shrug—and why he believes federal standards via NIST are the only realistic path. The discussion ranges from AI-driven healthcare and precision agriculture to robotics, demographics, labor culture, global supply chains, and what CES might look like in 2056.5 Key Takeaways from Gary:AI is now the spine of CES. CES 2026 centers on AI as infrastructure: CES Foundry at the Fontainebleau for AI + quantum, AI training tracks for strategy, implementation, agentic AI, and AI-driven marketing, and an AI-powered app helping attendees navigate the show.Fragmented state AI laws are an existential risk for startups. Over 1,200 state AI bills in 2025—including proposals to criminalize agentic AI counseling—could create a compliance maze only large incumbents can survive, which is why Gary argues for federal standards via NIST.Wearables are becoming systems, not gadgets. Oura rings, wrist devices, body sensors, and subdermal glucose monitors are starting to be designed as interoperable families of devices, with partnerships emerging to combine data into unified health services.Robotics is breaking out of the industrial niche. CES will showcase the largest robotics presence yet, moving beyond factory arms and drones to humanoids, logistics, social companions, and applied AI systems across sectors.Tariffs, alliances, and AI will reshape manufacturing. Gary is skeptical of “Fortress USA” strategies that try to onshore everything, pointing instead to allied reshoring (Latin America, Europe, Japan, South Korea) and the long-term role of AI-powered robotics in changing labor economics and global supply chains.This episode is brought to you by Zappar, creators of Mattercraft—the leading visual development environment for building immersive 3D web experiences for mobile headsets and desktop. Mattercraft combines the power of a game engine with the flexibility of the web, and now features an AI assistant that helps you design, code, and debug in real time, right in your browser. Whether you’re a developer, designer, or just getting started, start building smarter at mattercraft.io.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Caitlin Krause, author of Digital Wellbeing, argues that intentional design unlocks genuine connection within virtual spaces. Drawing on her teaching at Stanford and the University of Oregon, she's explored how XR environments can foster asynchronous connection and ambient awareness for people who crave belonging without hyper-social performance. Her framework rejects the "digital detox" model entirely—instead advocating for dignity-first design where users match attention with authentic intention.The hosts debate the deeper question: what happens to human purpose when AI handles all labor? Rony Abovitz frames this as the "asymmetry of design"—it's easy to build addictive tech, hard to build wellbeing tech. Caitlin counters that we may return to the original meaning of "amateur" (from amor, "to love"), where humans find meaning through play, creativity, and what Harvard's lifespan study confirms: quality of relationship and presence. The conversation spirals from platform ethics to post-work society to what first principles we should use when designing XR.5 Key Takeaways from Caitlin:Loneliness is a biological prompt to find another human—not a void to fill with endless content. XR can foster genuine forms of connection without requiring hyper-social performance.Dignity-first design unlocks freedom, invention, and agency. When digital spaces prioritize user agency over engagement metrics, people report feeling like they "got their life back."Science will soon prove what we already know about fractal patterns in nature and digital signals. The key is designing digital experiences that resonate with how humans biologically thrive.The "middle path" between nature and digital is both/and. Gamers building entire lives in virtual worlds can be healthy when those worlds offer creativity, belonging, and meaningful challenge.The post-labor economy needs a reset in literacy and values. When AI outperforms human workers, purpose shifts from survival to what makes you feel alive—maker culture, digital fab labs, hands-on creation, and "amateur" pursuits driven by love.In the News: Oracle, Silver Lake, and Abu Dhabi's MGX close the $50 billion TikTok spin-off deal. Meta cuts Reality Labs by 30%, but CTO Andrew Bosworth says it's moving to AI. The TCL glasses demo 70 grams of lighter, more advanced XR hardware than Ray-Ban Meta—proving that smart spending beats mega-spend.This episode is brought to you by Zappar, creators of Mattercraft—the leading visual development environment for building immersive 3D web experiences for mobile headsets and desktop. Build smarter at mattercraft.io.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Vince Kadlubek, co-founder of Meow Wolf, joins Charlie and Ted for a deep dive into the future of immersive entertainment, arguing that in an age of infinite AI-generated digital content, "physical reality is the only place novelty still exists." From Meow Wolf’s origins as a scrappy art collective dumpster-diving for materials in Santa Fe to becoming a global location-based entertainment juggernaut with new sites planned for Los Angeles and New York, Vince reveals the philosophy behind building "maximalist" worlds that don't just tell stories but allow audiences to inhabit them.In the news segment, Charlie and Ted discuss Netflix's $83B acquisition of Warner Bros (HBO/IP assets only), Meta cutting 30% of Reality Labs to fund AI while poaching Apple's top designer, and the looming battle for 2026 as Android XR prepares to launch.Vince breaks down Meow Wolf's evolution from static walkthrough experiences to "animated spatial storytelling" where environments and characters respond to user actions—a vision of "XR RPGs" (Extended Reality Role Playing Games) that bridge the gap between video games and theme parks. He explains why the "monoculture" of Game of Thrones is gone forever, why Netflix's acquisition power signals the end of traditional scarcity models, and why the future of storytelling isn't on a screen—it's cross-reality, persistent, and physically grounded.Guest HighlightsOrigins of the Multiverse: How a Santa Fe art collective turned a bowling alley into the "House of Eternal Return" with George R.R. Martin as landlord.The "Cross-Reality" Future: Why physical locations alone aren't enough—Meow Wolf is building a "mechanically connected transmedia universe" where your actions in the park affect your digital profile and vice versa.Hollywood 2.0: New LA location takes over a movie theater to "honor cinema" while deconstructing it into spatial storytelling.Novelty Theory: "I don't care about photorealistic AI gorillas anymore." Why digital content has zero value and physical presence is the new premium.Questing & Agency: New "XR RPG" mechanics in Dallas/Houston allow visitors to level up, solve puzzles, and impact the world—gamifying reality without headsets.News HighlightsNetflix acquires Warner Bros assets ($83B)—Streaming wars end with tech giants vacuuming up legacy IP; theaters face the "nail in the coffin."Meta cuts 30% of Reality Labs—Pivot to AI funding while hiring Apple's former design chief signals a shift from brute-force VR to refined wearables.Android XR & Samsung 2026—Google and Samsung prepare to challenge Vision Pro with a new ecosystem launch next year.Alibaba launches Quark AI Glasses—China enters the smart glasses race with multimodal AI assistants.Subscribe for weekly insider perspectives from veterans who aren't afraid to challenge Big Tech. New episodes every Tuesday. Watch full episodes on YouTube. Thanks to our sponsor Zappar!See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.