Another Crazy Day in AI: DeepMind Creates Interactive Virtual Worlds
- Wowza Team
- Aug 6
- 4 min read
Hello, AI Enthusiasts.
Ready to walk through your imagination?
Genie 3 just dropped from DeepMind. It builds entire worlds from text prompts—and this time, you can explore them in motion. Yes, that includes volcanoes and surreal creature zones.
Meanwhile, tech companies are spending AI money like it’s endless, outpacing what the U.S. allocates for education, jobs, and social services combined. And they’re not even close to done.
Over in the skies, Delta is letting AI set your ticket price in real time. No identity-based pricing, they say... but the algorithm definitely knows when you're ready to splurge.
Here's another crazy day in AI:
Google's Genie 3 creates interactive worlds from text
Silicon Valley pours $155B into AI, plans for hundreds more
Delta launches AI-powered ticket pricing system
Some AI tools to try out
TODAY'S FEATURED ITEM: New Model Builds Navigable Digital Spaces
Image Credit: Wowza (created with Ideogram)
What if you could instantly create any world you can imagine and explore it in real-time?
Google DeepMind has introduced Genie 3, a new world model designed to generate dynamic, interactive environments from text prompts. The model allows users to explore richly detailed simulated worlds in real time, at 24 frames per second and 720p resolution, while maintaining consistency in physical properties like terrain, lighting, and object placement for several minutes at a time. Jack Parker-Holder (Research Scientist) and Shlomi Fruchter (Research Director), along with other Google Researchers, led the development of Genie 3 as part of DeepMind’s continued work in world modeling and simulation-driven learning. Genie 3 builds on DeepMind’s earlier efforts in simulated environments, but introduces something new: not just realism, but responsiveness, and not just motion, but memory.
Unlike traditional generative tools that produce static images or videos, Genie 3 creates spaces that respond to movement and change. These environments can range from realistic natural scenes like glacial forests or volcanic landscapes, to imaginative settings filled with animated creatures or surreal geography. It also supports what DeepMind calls “promptable world events,” where additional text input can modify the scene in real time, triggering weather changes, introducing new elements, or adjusting the surroundings. This added layer of control makes the system useful not just for creative applications, but also for training and observing autonomous agents in evolving, simulated conditions.
What the model offers so far:
Generates immersive 3D environments directly from text
Enables real-time interaction at 24 fps with visual continuity
Maintains environmental coherence across minutes of activity
Allows mid-session text prompts to dynamically alter the environment
Supports simulation of both natural and fictional settings
Works with embodied agents to test behaviors and decision-making
Operates without requiring pre-built 3D assets or explicit geometry
Though Genie 3 introduces new possibilities for simulation, it remains an experimental system with clear limitations. The duration of interaction is currently short, and the range of user or agent actions is still somewhat limited. Simulating complex social interactions or replicating real-world geography with precision continues to be an area under development.
Still, the broader research direction here raises important questions. What does it mean for machines to move through space in ways that reflect our physical world, or entirely imagined ones? And how might this kind of model influence how people build, test, or experience environments, especially when those environments are designed to respond, not just exist? Genie 3 may not answer those questions outright, but it offers a glimpse of where world modeling could go next—not as a finished tool, but as an evolving area of research.
Read the full article here.
OTHER INTERESTING AI HIGHLIGHTS:
Silicon Valley Pours $155B into AI, Plans for Hundreds More
/Blake Montgomery, (Tech Editor), on The Guardian
Tech giants have already poured $155 billion into AI development this year—outpacing U.S. government spending on education, jobs, and social services. Financial disclosures from Meta, Amazon, Alphabet, and Microsoft reveal they’re nowhere near done, with combined capital expenditure expected to exceed $400 billion by year-end. Most of this spending is fueling the physical backbone of AI—servers, chips, and massive data centers. Despite the staggering figures, investors are thrilled, pushing company valuations even higher as the AI arms race accelerates.
Read more here.
Delta Launches AI-Powered Ticket Pricing System
/Ali Rogin (Correspondent) and Andrew Corkery (National Affairs Producer), on PBS News
Delta has become the first major airline to use AI to dynamically adjust ticket prices in real time, absorbing huge volumes of market and behavioral data to stay competitive. The airline insists it doesn’t personalize fares based on user identity, but concerns are growing about price discrimination and lack of transparency. With little regulation currently in place, some lawmakers and competitors are calling for guardrails to protect consumers. While other airlines like American remain skeptical, Delta’s move could shape the future of airline pricing.
Read more here.
SOME AI TOOLS TO TRY OUT:
That’s a wrap on today’s Almost Daily craziness.
Catch us almost every day—almost! 😉
EXCITING NEWS:
The Another Crazy Day in AI newsletter is on LinkedIn!!!
Leveraging AI for Enhanced Content: As part of our commitment to exploring new technologies, we used AI to help curate and refine our newsletters. This enriches our content and keeps us at the forefront of digital innovation, ensuring you stay informed with the latest trends and developments.
Comments