Enjoying this newsletter? Know someone who would too? Forward this email to them!
Sign up (it’s free)

I’m trying something new next week. On Tuesday you’ll get your first long-form deep dive, starting with Robotaxis - how they work, what the state of play is today, what still needs to happen for them to become boringly normal, and which companies could capture the most value of the multi-trillion dollar market. It’s designed for readers who want to go beyond headlines and really understand the frontier technologies that will reshape our world. Keep an eye on your inbox!

🇺🇸 US researchers boost concrete battery capacity 10×, enabling buildings to store renewable energy

MIT scientists have developed a new form of energy-storing concrete, ec³, that can hold 10 times more energy than previous versions, potentially transforming walls, roads, and bridges into large-scale batteries. The material uses cement, carbon black, and electrolytes to form a conductive nanonetwork, with organic electrolyte blends enabling up to 2 kWh of storage per cubic meter. Achieved by MIT’s EC³ Hub, the breakthrough could help stabilize renewable energy grids by integrating power storage directly into everyday infrastructure.

🇺🇸 🇫🇷 US and French innovators develop recycled nuclear batteries to power NASA's lunar missions

Zeno Power (US) and Orano (France) are collaborating to fuel space batteries with americium-241, a long-lived isotope recovered from recycled nuclear waste. These radioisotope power systems (RPSs) will provide decades-long, reliable energy for NASA's lunar rovers, landers, and infrastructure - critical for surviving the Moon’s frigid nights and shadowed regions. Detailed in a company announcement, the innovation offers a sustainable alternative to scarce plutonium-238 and supports NASA’s Artemis and Moon to Mars programs.

🇨🇳 Chinese researchers launch national robot training centers to accelerate humanoid AI development

China is building large-scale robot "boot camps" in cities like Beijing and Shanghai to train next-generation humanoid robots in real-world scenarios such as retail, elder care, and smart homes. Led by the government’s strategic focus on “embodied intelligence,” these centers will collect massive, standardized datasets, over 6 million data points annually from just one facility, to improve robotic AI and address current bottlenecks in data quality and scale. The initiative supports China's broader AI goals and positions it as a global rival to the US in humanoid robotics.

🇺🇸 US company NVIDIA equips humanoid robots with smarter brains and physics for real-world readiness

NVIDIA has launched major updates, including the Newton Physics Engine and Isaac GR00T N1.6 model, that give humanoid robots enhanced physical realism and reasoning capabilities. Developed with partners like Google DeepMind and Disney Research, Newton simulates complex tasks like walking on snow or handling fragile items, while GR00T integrates advanced language and vision reasoning through the Cosmos AI model. With this open, unified platform, NVIDIA aims to accelerate robot deployment from labs to homes and workplaces.

🇦🇺 Australian engineers unveil spider-like robot that 3D prints homes in a day, eyes lunar missions

Australian companies Crest Robotics and Earthbuilt Technology have developed “Charlotte,” a six-legged autonomous robot that can 3D print a 200 m² home in just 24 hours using local materials like sand and crushed brick. Designed to address housing shortages with low-cost, low-carbon construction, Charlotte’s compact, foldable design also makes it ideal for off-Earth building, including future Moon bases. Its Earthbag-inspired method and mobility offer a promising solution for rapid, sustainable infrastructure on both Earth and the lunar surface.

If you haven’t checked out our new partner Nanalyze yet, I highly recommend you do. This week they published a sharp breakdown on whether autonomous trucking is finally investable which you can read here, and released this video distilling 12 lessons from six once-popular companies they warned their readers to avoid - all of which went on to lose almost all their value.

🇵🇹 Portuguese researchers decode mice’s thoughts from facial expressions using machine learning

Scientists at the Champalimaud Foundation in Portugal have shown that machine learning can infer mice's internal thought strategies from subtle facial movements, offering a non-invasive alternative to monitoring brain activity. Published in Nature Neuroscience, the study found facial cues provided as much insight as neuron recordings and were consistent across animals - suggesting a universal link between thought and expression. While promising for brain research, the findings also raise ethical concerns about mental privacy as facial video analysis grows more powerful.

🇺🇸 US scientists create human eggs from skin cells, offering new hope for infertility treatment

Researchers at Oregon Health & Science University have developed a method to transform skin cells into egg cells capable of forming embryos, a breakthrough that could revolutionize infertility treatment. Using a process called “mitomeiosis,” the team created functional eggs without using stem cells, instead reprogramming skin cell nuclei inside donor eggs stripped of DNA. Published in Nature Communications, the study shows early promise for same-sex reproduction and treating egg loss, though clinical use remains years away.

🇩🇪 German researchers develop edible flu sensor that signals infection with thyme-like taste

Scientists at the University of Würzburg have created an edible molecular sensor that detects pre-symptomatic flu infection by releasing a thyme flavor in the mouth. The sensor combines thymol, a compound found in thyme, with influenza’s neuraminidase enzyme, which activates only in the presence of the virus, breaking bonds and triggering the taste. Published in ACS Central Science, this fast, low-cost approach could help curb flu spread in high-risk settings before symptoms emerge.

That’s all for today, please reply to this email if you have any comments or feedback, we’d love to hear from you about what we can do better!

Have you enjoyed this email? Make sure to share it with your friends and colleagues.

See you soon,

Max

Subscribe (it’s free)