Toggle light / dark theme

Get the latest international news and world events from around the world.

Log in for authorized contributors

3: Our Most Powerful Foundation World Model

Odyssey-3: a new step toward world models for physical AI

Odyssey has unveiled Odyssey-3, its latest foundation world model, designed to learn how objects move, interact, and respond to actions over time. Unlike conventional video generation that primarily produces visual sequences, Odyssey-3 aims to generate interactive environments that evolve in response to human or AI actions, potentially providing a more useful foundation for physical AI.

Built as an autoregressive diffusion transformer, the model learns patterns of physics, dynamics, and cause-and-effect from video, annotated events, gameplay, and simulated physical interactions. Its applications include generating real-time environments, creating training grounds for AI agents, and adapting learned representations to control physical systems.

Odyssey reports that Odyssey-3 Pro achieved a score of 66.1 on the Physics-IQ Verified video-to-video benchmark, which the company describes as a state-of-the-art result. Its evaluations on WorldMark also placed it first in three of four environment categories. In demonstrations, policies built using Odyssey-3 have been applied to robot-arm manipulation, humanoid tasks, and autonomous driving. The company reports that its driving policy was trained using just 20 hours of driving data while keeping the model’s backbone frozen.

The broader ambition is to move world models beyond visual prediction toward systems that can help machines anticipate how their environments change and learn how to act within them. If these capabilities generalize reliably beyond demonstrations and benchmarks, world models could become useful infrastructure for robotics, autonomous systems, and training increasingly capable AI agents.

The important caveat: generating physically plausible video is not the same as possessing a complete or accurate model of the real world. Benchmark results and demonstrations are promising, but robust transfer to unfamiliar conditions, reliable long-horizon predictions, and safe real-world control still require independent validation.

#worldmodels #robotics #AutonomousSystems #ArtificialIntelligence

This Tool Turns Scientific Articles Into Agents That Answer Questions and Collaborate

In other words, Paper2Agent is like a translator between a human-written paper and a chatbot.

That might sound redundant. After all, it’s already possible to upload a paper to ChatGPT, Claude, or another chatbot and ask questions. The difference is in the training: A chatbot can summarize a paper’s results, but it doesn’t have a deeper understanding of how those results came to be or whether the underlying analysis holds up.

By trying to replicate results based on the paper, Paper2Agent’s AI gets a sort of hands-on experience, potentially making it less prone to hallucination. The agents can “provide much more in-depth insights to the readers,” Zou told Nature.

The story of mycodiesel

Recently, a number of endophytic fungi have been discovered that produce volatile organic compounds (VOCs) whilst growing on agricultural waste substrates, whose chemistry is best defined as hydrocarbon and hydrocarbon-like. These compounds have potential use as both ‘green chemicals’ and fuels. This report discusses the discovery of the first fungus proposed as a producer of ‘Mycodiesel’. Also mentioned are many examples of fungi making these VOCs and some of the novel methods that have been specifically developed and used to study the fungal production of hydrocarbons. Finally, the report concludes with a discussion of commercial scale up and feasibility of this approach in helping to solve the world’s need for liquid fuels.

Exploring the Role of Industrial Hemp in Sustainable Aviation

The aviation industry faces increasing pressure to mitigate its environmental impact while maintaining operational efficiency. Industrial hemp, a versatile and sustainable crop, offers transformative potential across multiple facets of aviation: sustainable aviation fuel (SAF), carbon-neutral manufacturing materials, and carbon offset programs. Hemp’s ability to grow rapidly, its high yield of usable biomass, and its adaptability to marginal lands make it an excellent candidate for reducing aviation’s carbon footprint. This paper explores the technical specifications, economic feasibility, and environmental implications of incorporating industrial hemp into aviation, presenting a comprehensive case for its adoption.

Hydrogen-powered ‘rocket car’ tops 400 mph in record-breaking test (video)

That’s the fastest that any car with a hydrogen internal combustion engine has traveled. “It was an amazing experience,” Green told Space.com on Aug. 12 about his record-breaking drives.

“Anything above about 350 miles an hour [563 kph], we’d have gone away really pleased,” he added. “It is performing better. It was more reliable than we expected, and it has blown away the hydrogen record. It’s in every way a real thrill for all of us yesterday to see that happen.”

Green is no stranger to breaking records. In 1997, he drove a jet-powered car 763.035 mph (1,227.985 kph, or Mach 1.02), setting a new land speed record while also becoming the first person to break the sound barrier on land.

Hemp-Derived Graphene-like Materials: A Renewable Pathway Toward Scalable Conductive Carbon Nanomaterials

The scalable and sustainable production of graphene remains a significant challenge due to the high cost, complex processing, and environmental impact associated with fossil-derived graphite precursors. In this work, we report a biorenewable pathway for producing graphitic carbon from industrial hemp biomass, yielding a plant-derived material called CleanGraphene. This approach provides a renewable and potentially scalable alternative to petroleum- and coal-based graphene production while maintaining competitive structural and electrical performance. CleanGraphene samples are systematically characterized using X-ray diffraction (XRD), Raman spectroscopy, transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and thermogravimetric analysis (TGA) to evaluate crystallographic order, layer stacking, defect density, surface chemistry, and thermal stability. The results show that optimized CleanGraphene materials consist of multilayer graphene-like platelets with compact interlayer spacing (d(002) ≈ 3.36–3.37 Å), extended crystallite coherence lengths (Lc up to ~75 nm), large in-plane sp2 domains (La exceeding ~200 nm), and relatively low defect densities, indicating well-developed graphitic ordering. Electrical conductivity measurements using a binder-free pelletization method and four-point probe analysis demonstrate that the highest quality CleanGraphene samples achieve conductivities of (8.4–8.6) × 104 S m−1, surpassing leading commercial graphene benchmarks measured under identical conditions. Structure–property correlations confirm that electrical performance is governed primarily by crystallite coherence, defect density, and interlayer stacking order, while surface oxygen content plays a secondary role within an ordered graphitic framework. All CleanGraphene samples exhibit excellent thermal stability, retaining more than 95% mass up to ~800–900 °C under an inert atmosphere. Collectively, these findings establish quantitative quality benchmarks for hemp-derived graphene and demonstrate that biomass-based graphene can achieve electrical and thermal performance comparable to, and in some cases exceeding, conventional commercial products. This work highlights industrial hemp as a promising renewable precursor for the scalable production of high-performance graphitic nanomaterials for electrically and thermally conductive composite applications.

Electricity-generating bacteria may power future innovations

A team led by Rice University bioscientist Caroline Ajo-Franklin has discovered how certain bacteria breathe by generating electricity, using a natural process that pushes electrons into their surroundings instead of breathing on oxygen. The findings, published in Cell last month, could enable new developments in clean energy and industrial biotechnology.

By identifying how these bacteria expel electrons externally, the researchers offer a glimpse into a previously hidden strategy of bacterial life. This work, which merges biology with electrochemistry, lays the groundwork for future technologies that harness the unique capabilities of these microscopic organisms.

“Our research not only solves a long-standing scientific mystery, but it also points to a new and potentially widespread survival strategy in nature,” said Ajo-Franklin, professor of biosciences, director of the Rice Synthetic Biology Institute and a Cancer Prevention and Research Institute of Texas (CPRIT) Scholar.

/* */