The AI XR Podcast

Seventeen years into building the world's largest XR conference, Ori Inbar is not prone to hyperbole. He has watched hype cycles inflate and collapse, made predictions that turned out too optimistic, and learned to hold claims carefully. That is what makes his framing of AWE 2026 worth paying attention to: he calls it the most consequential year in the show's history. Not because everything is working — there have been heartbreaking layoffs in some corners of the industry — but because the convergence happening right now between AI and spatial computing is unlike anything the field has seen before.
Before Ori joins, the hosts wade through a week of signal and noise. Three big IPOs — Cerebras, Quantium, and others — are absorbing investor attention, with Quantium carrying a $15 billion market cap on what Charlie calls "de minimis revenue," raising questions about whether the quantum AI bandwagon has lapped actual quantum utility. Rony poses the challenge directly: what is the real use case for quantum computing besides breaking encryption?
Ted and Rony unpack qubits, superposition, and state-based computing in plain language, with Rony floating the multiverse analogy and the idea that neurons themselves might function as tiny quantum computers. Ted points to Cleveland Clinic's work with IBM as a concrete near-term example. Elsewhere: Anthropic has announced it is close to recursive self-improving AI — which, as Charlie notes, is the textbook definition of AGI. Apple's WWDC is around the corner, promising a Siri overhaul, a folding iPhone, and — per a Rony tease — a secret wearable project he can say almost nothing about.
Google's new Dream Beans app indexes all your files and reads your emails for $70 extra on top of YouTube Premium; Ted cuts through it simply: "It's an ad play." Suno is raising $400 million at a $5.4 billion valuation while still being sued for training on copyrighted music — Rony's investor thesis frames it as AI-generated sound engineered to trigger neurologic adrenaline hits, an addiction feedback loop he calls "musical crack." And Fox is reformatting "Farmer Wants a Wife" into 101 vertical two-minute episodes — Charlie's read is that this may be the moment when video content permanently shifts from advertising-supported to direct-to-consumer.
When Ori arrives, the conversation opens on Snap. Evan Spiegel is expected to make a major consumer announcement at AWE — Ori says Snap has put all their eggs in this basket, and the audience at the show will be the first to see it. Ted frames the stakes plainly: if the price shocks people, it's a consumer breakthrough; if it's expensive and exotic, it stays in the science column. Snap recently acquired Illumix, a spatial universe understanding startup, a move that signals the company is building seriously in this space.
More broadly, Meta's retreat from the metaverse has acted as a talent accelerant — hundreds of experienced XR engineers are now building independently, and the resulting Cambrian explosion of content for both AI glasses and AR glasses is just beginning. Ori points to Supernatural spinning off from Meta at a fraction of what the company paid for it as a symbol of a broader reset: the talent is free, the tools are ready, and the content is coming.
The tools argument extends to development itself. Ori built a working AR prototype — he describes it as a "chat with animals" experience — in two days using Gemini in the backend. Vibe coding for spatial computing is no longer theoretical. Unity has added it. Google is building game worlds by prompting them into existence. Ted's broader thesis: AI has learned from game engines well enough that it may not need them for 90% of use cases — a provocative observation for an industry built on Unity and Unreal.
AWE's show floor this year reflects how much is happening at once: 250 exhibitors, a dedicated smart glasses pavilion, 25 studios and LBEs launching new games, the first-ever art festival with a single juried winner, research paper posters bridging academia and industry for the first time, a digital twins pavilion, and 300 enterprise attendees expected. Ori calls it "almost 10 festivals in one." The iSpatial theme — deliberately constructed as a counter to iRobot — puts humans at the center of spatial AI. Ori was using the phrase "spatial AI" years before the current AI wave hit; he had called it "SPAI" in a pandemic-era keynote. His three biggest XR trends of 2026: AI smart glasses, AI-generated content for glasses, and world models colliding with XR.
The endgame vision comes from Rony: Oakley-weight wraparound glasses at 30–40 grams, human retina resolution, full indoor/outdoor capability, AR and VR combined, wireless, all variable focus, under $500. Ted adds that it also has to land under $650 fully costed at retail. Ori's honest answer: "I promised myself I'm not gonna predict when this happens. I've tried many times and was always way too optimistic." Ted teases Gixel, a German startup he and Rony are involved in using non-waveguide display technology already above 60 pixels per degree — when you put the prototype on, he says, it is crystal clear.
Defense is the fastest-growing vertical at AWE. Healthcare, manufacturing, aerospace, and automotive are major enterprise sectors. Digital twins are the biggest thing in enterprise XR right now, with world models emerging as the intelligence layer sitting beneath them. Over 10 million AI glasses — display-free — sold last year. Ori's framing of why display glasses matter more: AI is shifting these devices from tools that help you learn about things to tools that actually do things.
Key moments:
[00:02:47] Quantum IPO bubble — Rony asks what the actual use case is
[00:05:48] Quantum mechanics in plain language — qubits, superposition, neurons as quantum computers
[00:09:51] Apple WWDC preview — Siri, folding phone, Rony's secret Apple wearable tease
[00:11:38] Google Dream Beans — Ted: "It's an ad play"
[00:12:51] Suno $400M raise — Rony: "Musical crack" and the TikTok-for-music thesis
[00:14:42] Fox reformats "Farmer Wants a Wife" into 101 vertical episodes — the content inflection point
[00:17:00] Ori joins — AWE 2026 as "most consequential year in our history"
[00:17:40] Snap and Evan Spiegel's expected consumer announcement at AWE
[00:19:38] Cambrian explosion of XR content — Meta talent diaspora, Supernatural spinoff
[00:23:07] Vibe coding for XR — Ori's AR prototype built in two days with Gemini
[00:25:48] Charlie inducted into the AWE Hall of Fame — joining Ted and Rony
[00:28:36] iSpatial theme — Ori's three biggest XR trends: AI glasses, AI content, world models
[00:39:31] Defense fastest-growing vertical. Digital twins biggest in enterprise.
[00:47:18] Rony's endgame AR glasses vision. Ted teases Gixel's crystal-clear prototype.

So what? The XR field has spent years promising a future that kept receding. What this conversation makes plain is that the receding has stopped. The content is coming, the hardware is multiplying, the development tools have dropped the floor on what it takes to build, and the enterprise use cases are no longer pilot projects.
AWE 2026 is a snapshot of something already underway.
Brought to you by Zappar and Mattercraft. Build web-based AR experiences without writing code at mattercraft.io.
Find full episodes where you listen to podcasts and watch on YouTube https://youtu.be/MUJW1Kp0Yu4
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

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 2026

So 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/txUicSGoe98

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Charlie Fink, Ted Schilowitz, and Rony Abovitz take the full hour to work through the most consequential AI and spatial computing stories of the moment — unfiltered, in depth, and without the usual polite hedging that comes with having someone on to promote something. This is a pure news and commentary episode, and the news is strange enough that three experienced people sitting in a room still cannot fully account for it.
AI XR News You Should Know:
The OpenAI vs. Elon Musk case concluded without a clear ruling, but the more durable observation is what the whole saga revealed about Sam Altman. He has now survived being ousted by his own board (which he subsequently dismantled), a high-profile lawsuit from Elon Musk, and senior rivals leaving for government roles. Rony frames this through the Overton window — Altman studies what society is prepared to accept at any given moment and positions himself precisely there. Ted references a New Yorker profile that describes Altman as having a politician's gift for telling people what they want to hear until it becomes true. The question hanging over that observation: are the large language models themselves doing the same thing, reflecting the values and methods of the people who built them?
The financial architecture underneath the AI boom looks precarious on close inspection. SpaceX, widely assumed to be profitable, is losing five billion dollars a year. Anthropic is spending three dollars for every dollar of revenue it generates — and is paying SpaceX approximately one billion dollars a month for compute through roughly 2030. Rony's framing lands hard: two money-losing entities are funding each other while NVIDIA captures all the margin in between. Sequoia published a fifty-page analysis arguing the economics cannot work — while simultaneously holding positions in the companies it is critiquing. Charlie calls it a Ponzi scheme for venture money. Ted's counterpoint: the company that cracks the ad-supported free tier wins everything, and Google already knows how to do that.
Google I/O delivered less on wearables than expected, but the real story was a deliberate strategic decision to put Gemini at the center of the company's entire product surface — effectively cannibalizing an eighty-two-billion-dollar search business before a competitor does it for them. The Innovator's Dilemma, run on purpose. On the hardware side, Android XR glasses are designed to be imperceptible as technology — thin temples, hidden camera portals, frames that belong in an optometrist's display case rather than a trade show floor. Rony notes that Google's glasses almost certainly incorporate Magic Leap optics, following a partnership announced in fall 2025. The accountability conversation runs through all of it: Ted's observation that AI companies are not building airplanes — and therefore carry none of the legal liability that comes with critical infrastructure — gives shape to a broader anxiety about who is keeping the ledger on what these systems actually do to people.
[00:00] – Cold open and episode framing: why there is no guest today and what the trio plans to cover.
[04:15] – OpenAI vs. Elon Musk non-verdict: what the outcome (and lack of one) actually reveals.
[09:30] – Sam Altman and the Overton window: Rony's read on how Altman has survived everything thrown at him.
[16:00] – Anti-AI backlash on campuses: Eric Schmidt booed at University of Arizona, YouGov poll showing 69 percent of young people negative on AI, and what the demographic gradient means.
[24:45] – SpaceX financials and the AI funding loop: the five-billion-dollar annual loss, Anthropic's burn rate, and Charlie's Ponzi scheme framing.
[33:20] – Sequoia's fifty-page report and the ad model endgame: Ted's argument that Google wins because they already know the business model.
[41:00] – Google I/O: the deliberate destruction of the search business, Android XR glasses, and why distribution beats specifications.
[49:10] – AI accountability and the airplane analogy: Ted's line, Rony's "underground noise" from generals and CTOs, and the problem of regulatory vocabulary.
[55:30] – Palantir, dual-use opacity, and the Lookout Mountain Air Force Station story: Rony on Jared Leto, classified film studios, and Cold War bunkers in Laurel Canyon.
[01:01:00] – The success ledger: who is measuring impact, and what should actually count as winning.

So what? The through line connecting every story in this episode is the question of accountability — who is keeping track, who has the standing and vocabulary to push back, and what happens when the answer to both questions is nobody. Sam Altman's resilience, the AI funding loop, Google's search gamble, the campus backlash, the airplane analogy — all of it points to an industry moving faster than the systems designed to check it. That may be fine. It may not be.

Charlie, Ted, and Rony do not pretend to have the answer, but they ask the question harder than most.
This episode is sponsored by Zappar and Mattercraft. Mattercraft is Zappar's web-based platform for building augmented reality experiences without an app. Find them at mattercraft.io.
Two free AWE tickets are still available — a nine-hundred-dollar value each. Reach out to Charlie Fink on socials.

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Joshua Pantony spent years being told there would never be a viable AI company in his lifetime. He sold his first AI company to Microsoft anyway — work that quietly became part of what is now Microsoft Copilot. Today he runs Boosted AI, an agentic platform serving more than 400 institutional investors who collectively manage around five trillion dollars in assets. He is one of the most credible voices in applied AI finance, and his read on where the industry is heading cuts through a lot of noise.
The conversation covers what it actually means to deploy AI in professional investing — not the demo version, but the one that has to earn trust from portfolio managers who have built careers on discretion and judgment. The platform learns each investor's individual style and then acts like a highly motivated junior analyst who never sleeps: constantly surfacing ideas, flagging risks, and improving the workflow without ever taking over the decision.
Josh also unpacks why the Bloomberg terminal is facing its BlackBerry moment, why the technology moat is effectively dead, and why the next durable advantage in finance will come from human trust networks that no model can replicate. Ted Schilowitz and Rony Abovitz join host Charlie Fink with sharp frames throughout — Rony's observation that the innovator's dilemma is now a high-frequency problem landed hard.
AI XR News You Should Know: The episode opens with two news segments covering AWE 2026 and the Snap Spectacles keynote with Evan Spiegel, the Samsung Galaxy Glasses debut, Gemini rolling out as Android's native agentic AI, the Cerebras sixty-billion-dollar IPO, and what an AI filmmaking company launched by the creators of Instagram Stories tells us about the future of short-form content. The conversation about micro-dramas, why Quibi failed, and what sixty percent of social media users now say about their own feeds leads directly into the trust themes that run through the entire episode.
Key Moments:
[00:00] – Cold open and welcome. Charlie frames the sixth anniversary of the show.
[02:30] – AWE 2026 recap. Snap Spectacles keynote, Evan Spiegel on stage, Samsung Galaxy Glasses previewed.
[06:00] – Gemini as Android's native agentic layer. What it means that AI is now replacing the OS interface.
[09:15] – Cerebras sixty-billion-dollar IPO. What a big AI IPO year signals for the sector.
[12:00] – AI filmmaking and Instagram Stories creators. The new short-form production economy.
[14:30] – Why Quibi really failed. No sharing mechanic, wrong bet on clipping, and arriving before the audience was ready.
[16:45] – The trust problem in social feeds. YouGov data: sixty percent of users cannot tell what is real. Social becoming a lie stream.
[19:00] – Guest intro. Joshua Pantony on being told AI would never be a viable business, and the algorithm he wrote at twenty that saved a million dollars.
[24:00] – How Boosted AI works. The digital twin model, the agentic workflow, and why it is not a portfolio manager.
[33:00] – The Bloomberg terminal's BlackBerry moment. Thirty thousand dollars a year for what AI will deliver for a fraction.
[42:00] – The moat is dead. Why user context — not the technology — is the durable advantage.
[51:00] – The innovator's dilemma at high frequency. Rony on why a day in AI is like a decade, and what that means for incumbents.
[58:00] – Trust networks as the last edge. The analog handshake as the most valuable currency in a world of synthetic information.

This conversation is a clear-eyed look at what it takes to build AI that professionals actually adopt — not a pitch, not a thought experiment. Josh's framing of Wall Street as the greatest collective intelligence humanity has built, and his argument that AI can finally make capital allocation genuinely more efficient, gives the episode an ambition that goes well beyond fintech.
The question of what survives automation — and what only humans can do — runs underneath every answer.
This episode is sponsored by Zappar and Mattercraft. Mattercraft is Zappar's browser-based augmented reality creation platform — build and deploy WebAR experiences without an app, at mattercraft.io.

If you like what you hear, subscribe to The AI XR Podcast on Apple Podcasts or wherever you listen. Watch on YouTube - https://youtu.be/I8hLgBneUas

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

As director of Keyhole, Dave Lorenzini delivered the 3D Earth zooms that ran on CNN during the 2003 Iraq War — netting five million users in a month. Sergey Brin was one of them. Google bought the company and poured in billions to build, fuel, and serve maps. As Google Earth, it forever changed how we relate to space.
From there: pioneering work on Google Glass, AR platforms, and running an immersive XR lab in Europe for Draw & Code exploring the future of spaces, places, and faces. Today Dave directs Quantum Studio, building World Agent and 4D ID — the "DNS for real space," an addressing layer where every place, object, and moment gets a name AI systems can agree on. His thesis: the next decade of AI won't run on better maps. AI needs an operating system for reality. Not a map. Not a database. A living, queryable foundation where every place on Earth answers for itself.
AI XR News: The OpenAI vs. Musk trial continued with damaging testimony from Mira Murati and Greg Brockman. Anthropic struck an unholy alliance with xAI's Colossus compute. GameStop bid for eBay. Colin Angle is back with Familiar, an AI robot pet. Coinbase cut staff. Ask.com finally died. VRChat hit 100,000 concurrent daily users in Japan.
Key Moments:
[00:03:34] AWE Long Beach in 30 days: Dave on the board, Snap glasses expected, 400 speakers and 250 exhibitors
[00:20:10] 30 AI glasses coming: why the near future belongs to audio-first, AI-powered smart glasses
[00:25:34] Keyhole origin story: satellite imagery, $25K/sq mile, Sergey Brin, and a $500M/year acquisition
[00:37:30] Google Glass, Luxottica, and why Google blinked when it could have been 10 years ahead of Meta
[00:40:00] XR vs. rockets: why building for the human brain is harder than getting to Mars

Brought to you by Zappar, the company behind Mattercraft — the leading visual development environment for immersive 3D web experiences. Start building at mattercraft.io.

Subscribe to the AI XR Podcast wherever you listen to podcasts. Watch the full episode on YouTube: https://youtu.be/weNANIIo7EA

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.