Why world model startups are sitting on billions and telling no one anything. Tl;dr. * World model startups like World Labs, Odyssey and Decart have raised hundreds of millions on the promise of AI that understands 3D space and physics, but have revealed almost no technical details about how their systems actually work. * Founders and even their data partners are refusing interviews and staying vague on roadmaps to protect proprietary data deals, avoid tipping off Google DeepMind and Meta, and buy time while the tech remains unreliable. * That secrecy is slowing independent evaluation and raising fears of a hype bubble, legal fights over training data, and a future where spatial intelligence is locked behind a few closed APIs. The billions are real. Silicon Valley has not been this excited - or this confused - about a category since the first ChatGPT boom. World Labs, the startup founded by Stanford AI pioneer Fei-Fei Li, has raised more than $230 million at a valuation north of $1 billion from Andreessen Horowitz, NEA and Radical Ventures. Israeli startup Decart has pulled in over $150 million for its real-time interactive models. Odyssey, founded by former Cruise and Voyage leaders Oliver Cameron and Jeff Hawke, has raised more than $45 million from EQT Ventures, GV and Air Street Capital. Luma, Runway and a wave of younger labs are all rebranding around the same phrase: world models. The pitch is enormous. Not chatbots that predict the next word, but AI systems that generate persistent, navigable 3D worlds, understand physics, object permanence and spatial reasoning, and eventually power everything from video games and films to robots and self-driving cars. Investors call it the next step to artificial general intelligence. But ask what is actually under the hood, and the conversation stops. A demo, not A paper. Unlike the large language model era, where startups and labs competed to publish papers, release weights and top leaderboards, world model companies are shipping tightly controlled demos with almost no technical disclosure. World Labs released Marble, its browser-based text-to-3D environment tool, in June 2025, followed by a first large world model preview that turns a single image into an explorable 3D scene. Odyssey launched its Explorer tool for interactive streaming video worlds. Decart went viral with Oasis, a playable AI-generated Minecraft clone, and later a real-time conversational video model. DeepMind upped the pressure in August 2025 with Genie 3, a real-time interactive world generator. All are impressive in a browser tab. None came with architecture details, training compute numbers, evaluation benchmarks, or clear roadmaps for pricing, API access, or when these toys become reliable products. Requests for model cards, data sources and failure rates are met with waitlists and marketing language about spatial intelligence. The founders who won't talk. That silence is deliberate. In recent months, founders of three of the most-funded world model startups have declined in-depth technical interviews or pulled out of conference panels, even as they continue to raise and hire aggressively. Off the record, employees and investors give the same reasons: no one wants to show their hand. First, the race is unusually paranoid. With Google DeepMind, Meta, Nvidia and OpenAI's Sora team all building competing video-to-world systems, a single detail about data pipelines, rendering engines or hybrid diffusion-transformer designs could be copied in weeks. Second, the tech is still brittle. Anyone who has spent time in an AI-generated world knows the problem: walls melt, objects disappear when you turn around, avatars fall through floors. Founders prefer to talk about long-term vision - persistent worlds for creators and robots - rather than admit their current models can only hold coherence for a few minutes. Third, hype is the strategy. In a market where talent and GPU capacity are scarce, mystery helps. Being the stealthy Fei-Fei Li company or the secretive ex-Cruise team is better for recruiting and fundraising than publishing a benchmark where you lose to Genie 3. Even the data suppliers have gone quiet. More telling is who else won't talk: the data suppliers. World models need something LLMs never did - massive libraries of 3D geometry, video with camera movements, drone footage, game engine renders, and licensed scans of real-world places. Companies that provide that fuel, from stock video platforms and photogrammetry studios to game asset marketplaces and autonomous vehicle fleets, have signed exclusive deals with world labs but now refuse to name clients or discuss terms. Two data executives told reporters this summer they were barred by NDAs from confirming partnerships, citing fears of copyright lawsuits and backlash from creators who never consented to having their 3D work used for training. It echoes the early LLM data wars, but with higher stakes. A text scrape is one thing. A licensed 3D replica of a city, a national park or someone's Unreal Engine environment raises new questions about ownership, consent and price. No startup wants to be the test case, so no one discloses where the worlds come from. What secrecy means for the future. The short-term winner from secrecy is the incumbents. DeepMind, Meta's WorldGen efforts and Nvidia's Cosmos platform can afford to experiment in public because they already own data, distribution and compute. Startups, by contrast, are trying to build moats out of mystery. But researchers warn of three long-term costs. One is evaluation chaos. Without open benchmarks for physics consistency, memory and controllability, there is no way to compare World Labs vs. Odyssey vs. Decart vs. Genie 3 except vibes from cherry-picked demos. That makes enterprise adoption almost impossible for gaming studios and robotics firms that need reliability guarantees. Two is a legal time bomb. If training data deals stay hidden, Hollywood studios, game publishers and 3D artists say they cannot audit infringement or negotiate fair licensing. Lawsuits that hit image and video generators in 2023-2024 could return even larger for 3D worlds. Three is centralization. The dream of world models was open, creator-driven universes anyone could build. The current path points to the opposite: a handful of closed APIs where you rent explorable worlds by the minute, with pricing, limits and safety filters controlled entirely by the lab. Until one company breaks ranks and actually explains how its worlds are made - what data, what model, what roadmap - customers, researchers and even investors are left betting billions on faith. AndroGuider Team Articles written by the AndroGuider team. Androguider try to make them thorough and informational while being easy to read.

What AWS's chief AI and technology officer told employees about the future of AI. At the AWS Global Meeting, Matt Wood drew a parallel between AI today and cloud's early days - and brought a live demo to prove the future is already here. Published: September 17, 2026 Key takeaways. * AWS is building the foundational tools for AI the same way it built them for cloud computing. * Odyssey unveiled Agora-2, an AI that simulates shared worlds with no game engine. * From world models to code synthesis, frontier startups are building on AWS Trainium. At the AWS Global Meeting on September 17, AWS Chief AI and Technology Officer Matt Wood laid out a vision for where AI is heading - and brought a live demo on stage to prove it's already arriving. "What we are doing today at AWS for AI is exactly what we did for cloud computing, to make this technology broadly distributed and democratized for everybody," Wood told the audience. Then he named the building blocks: Trainium for compute, Bedrock for inference, AgentCore for agents - and brought the startups already building on them to prove it. University researchers are using Amazon's Trainium chip to push the boundaries of their work - and making the chip better for all developers in the process. A startup demonstrates its most realistic world simulator yet. The centerpiece of the meeting was a live demonstration from Odyssey, a startup building foundation world models - a new kind of AI that learns physics by watching massive amounts of video rather than relying on hand-coded rules. "A language model learns to predict the next most likely word in a sentence. A world model learns to predict the next most likely future of a world," explained Oliver Cameron, CEO of Odyssey. World models learn how the world works - from physics and cause-and-effect to human behavior - providing a foundational intelligence that can be applied across many industries. A world model draws on extensive learned knowledge of the world to automate physical systems, train AIs in interactive environments, and much more across robotics, autonomous driving, science, healthcare, education, gaming, defense, energy, and beyond. Through a new Student Rewards program on AWS Builder Center, university students worldwide can earn credits, premium training, and certification vouchers. On stage Cameron introduced Odyssey-3, Odyssey's most advanced foundation world model yet, which is trained and runs on Trainium3. Odyssey-3 is trained on vast observations of the world, giving it a learned grasp of physics, cause-and-effect, and human behavior. The same model can control robots, power humanoids, drive vehicles, pilot drones, train other AIs, and play video games. Because Odyssey-3 already understands how the world works, it can learn new tasks quickly, with far less task-specific data than previous systems. Additionally, Cameron unveiled Agora-2, a multi-agent world model - the first to hold multiple intelligences inside the same simulated world at the same time. Cameron and Wood each picked up controllers and entered Agora-2 in real time. The audience watched on a split screen as both navigated, interacted, and fought within the same AI-generated environment, each from their own perspective. Agora-2 supports up to 20 participants in a single simulation, with no game engine underneath. Cameron, who spent a decade working on autonomous vehicles before founding Odyssey, said the implications go well beyond gaming. "I believe we are on the cusp of a GPT-3 moment for world models," he said. "This is the moment where we see an explosion of applications for world models, and their adoption into every industry imaginable." OpenClaw creator Peter Steinberger on AI's next chapter. The meeting also featured a fireside chat between Peter Steinberger, the creator of OpenClaw and now at OpenAI, and AWS Vice President Francesca Vasquez. Steinberger discussed how the tools and infrastructure available to AI developers are evolving, a conversation that underscored Wood's broader argument that the biggest names in AI are building on and alongside AWS. Steinberger has become one of open source's biggest recent success stories. His OpenClaw project, which Vasquez called "the fastest growing open source project in GitHub history," has drawn more than 388,000 GitHub stars and 3.2 million active users. The bigger shift, he argued, is in how software itself gets built. "It's no longer 'I write code.' It's more, how can I design a system so my agent can do its best work?" As he put it, "We're all moving a little bit up the ladder. We're not just coders anymore." AWS is giving students at 132 universities across 18 countries a full year of free access to Kiro, so they enter the workforce ready to build with AI from day one. Why frontier AI startups are choosing AWS Trainium. A common thread across the meeting was the role of AWS Trainium in powering frontier AI workloads that demand sustained compute over weeks - a profile that general-purpose hardware struggles to sustain. Odyssey-3 runs on Trainium3 at over 80% model flops utilization - roughly double the industry average, according to the company. Odyssey joins a growing network of startups building on Trainium, including DeCart AI (real-time game environments), Neura Robotics (physical AI), Reactor (world model infrastructure), Splash Music (music generation), and Poolside (code synthesis). "We're here because you moved fast when we needed it, and then built something really special with us," Cameron said of AWS. Wood closed with a vision for where all of it is heading. "Our goal at AWS is to make it as easy, cheap, and fast to build these valuable economically viable AI systems as it is for anyone to build a website today," he said. "It's happening now, it's happening fast, and it's happening in all industries, but most importantly of all, it's happening here on AWS."

Odyssey has raised $310 million in Series B funding at a $1.45 billion valuation. Natural Capital led the round, with participation from Amazon, GV, AMD, EQT, and IQT. Existing investors include Jeff Dean, Elad Gil, Qasar Younis, Kyle Vogt, and Garry Tan, alongside researchers from OpenAI, DeepMind, MSL, Recursive, and Thinking Machines. The funds will accelerate the company's development of AI capable of understanding and simulating the world. Latham & Watkins advised AMD in the transaction.