Weekly Piece of Future #122
From Neural Implants to Quantum Breakthroughs and Non-Binary Chips
Hey there, fellow future-addicts!
Welcome to this week's edition of Rushing Robotics! This week’s newsletter is packed with groundbreaking advances—from brain implants that observe neural beginnings to AI models teaching machines to understand the physical world. Whether it’s biotech, robotics, or AI breakthroughs, we’ve got you covered with fresh insights and cutting-edge innovations shaping tomorrow.
🤯 Mind-Blowing
Discover how Harvard researchers developed a soft neural implant to study early brain development in tadpoles, unlocking secrets of neurodevelopmental disorders. Explore Meta’s V-JEPA 2, a video-trained AI model pushing machines closer to real-world physical reasoning. Plus, see how NHS England plans to make robotic surgery the gold standard by 2035, and marvel at Oxford’s record-smashing quantum computing accuracy. Meta’s new AI lab aiming for superintelligence signals the dawn of a new era in AI research.
🔊 Industry Insights & Updates
From Beijing’s RoboBrain 2.0 powering smarter humanoid robots to Wandercraft’s new industrial robot Calvin, the robotics landscape is evolving fast. Learn about Argonne National Lab’s AI-enhanced defect detection in 3D metal printing and China’s pioneering non-binary chips revolutionizing computing power. These global advances highlight the race toward smarter, more efficient machines.
🧬 BioTech
Australian researchers are pioneering mRNA tech to expose hidden HIV reservoirs, bringing us closer to a cure. The University of Virginia’s interactive artificial pancreas uses digital twins to personalize diabetes care. Meanwhile, Chinese scientists have developed magnetic microrobots that manipulate droplets with unprecedented speed and precision—opening doors to advanced medical and chemical applications.
💡 Products/Tools of the Week
Boost your productivity with Chat4Data, a Chrome extension that scrapes web data via conversational AI, and explore Skywork AI’s powerhouse workspace that deepens your research by tenfold. For developers, PureCode AI accelerates full-stack software delivery, while Tapflow helps experts monetize knowledge by turning it into digital courses and guides.
🎥 Video Section
Don’t miss the latest in robotics motion and AI demonstrations—featuring Wandercraft’s Calvin, Boston Dynamics’ Spot & Orbit in action, and KAIST’s agile humanoid leg performing flips and hops. These videos showcase how far robotics and AI have come—and hint at what’s next.
Innovation is moving faster than ever, and we’re here to keep you in the loop every step of the way. Stay hungry, stay futurish!
🤯 Mind-Blowing
Bioengineering experts in the U.S. have unveiled a soft and highly adaptable neural implant designed to study the earliest stages of brain development. Created at Harvard’s SEAS, the bioelectronic device was implanted into the neural plates of tadpole embryos, enabling detailed observation of neural activity as the nervous system begins to take shape. Hao Sheng, PhD, a postdoctoral fellow and lead author of the study, said the device integrated naturally with the brain’s developing structure, enabling precise neuronal activity recording without altering normal embryonic growth or behavior. This technology paves the way for more advanced models of early brain formation and holds significant promise for the study and prevention of neurodevelopmental disorders.
Meta has introduced V-JEPA 2, a cutting-edge AI model engineered to enhance machines’ ability to comprehend and anticipate physical interactions in the real world. Trained exclusively on video data, V-JEPA 2 significantly advances the capacity of robots and AI agents to reason about cause and effect, enabling predictive decision-making prior to action. The model departs from traditional approaches by simulating physical reality without relying on large datasets of labeled content. Operating within a latent space, it offers improved speed and adaptability. Meta views this as a vital milestone on the path to artificial general intelligence. The company also unveiled three new evaluation benchmarks to aid researchers in measuring AI’s reasoning and video learning capabilities. V-JEPA 2 is particularly suited for autonomous systems—such as delivery robots and self-driving vehicles—that require real-time interpretation of their environments and predictive planning.
The head of NHS England is set to announce that by 2035, robot-assisted surgery will become the standard approach for 90% of keyhole procedures. This initiative, aimed at significantly reducing the NHS hospital treatment backlog, will see a dramatic rise in the number of patients receiving robotic surgery—from 70,000 annually today to 500,000 by 2035. The expansion will impact treatments including cancer surgeries, hysterectomies, joint replacements, and emergency interventions. Robotic systems, either remotely operated by surgeons via 3D consoles or pre-programmed for specific tasks, offer greater precision and have been shown to speed up recovery and hospital discharge times. By 2035, robotic surgery is expected to be the default method for many minimally invasive operations.
Physicists at the University of Oxford have achieved the most precise control of a quantum bit (qubit) ever recorded, registering just one error in 6.7 million single-qubit operations—equivalent to an error rate of 0.000015 percent. This marks a nearly tenfold improvement over their own previous world record, set a decade ago. The results are set to be published in Physical Review Letters under the title “Single-qubit gates with errors at the 10⁻⁷ level.” To put the achievement into perspective, Oxford researchers note that an individual is more likely to be struck by lightning in a given year (1 in 1.2 million) than to witness a failure in one of their quantum gates. This dramatic leap in accuracy addresses a fundamental challenge in transitioning quantum computers from experimental setups to real-world applications.
Meta is preparing to launch a new artificial intelligence research lab with the ambitious goal of developing "superintelligence"—a theoretical form of AI that surpasses human cognitive abilities. According to four individuals familiar with the company's plans, this initiative is part of a broader restructuring of Meta’s AI strategy, now being led directly by CEO Mark Zuckerberg. The lab will include prominent figures in the field, including Alexandr Wang, founder of Scale AI. This move reflects Meta’s determination to remain at the forefront of AI innovation in an increasingly competitive global technology landscape.
🔊 Industry Insights & Updates
The Beijing Academy of Artificial Intelligence (BAAI), a prominent non-profit research institute, has unveiled RoboBrain 2.0—an open-source AI model engineered to serve as the cognitive core for China’s next generation of humanoid robots. Announced at BAAI’s annual conference in Beijing, the launch marks a significant stride in China’s broader push to enhance robotic intelligence through cutting-edge AI. RoboBrain 2.0 boasts enhanced spatial awareness, enabling robots to more precisely gauge distances from nearby objects, and advanced task planning that allows for autonomous breakdown of complex activities into simpler, accurate steps. According to BAAI head Wang Zhongyuan, the new model is the most powerful open-source AI framework globally for robotics, with the potential to accelerate the development and deployment of humanoid robots in China’s rapidly evolving robotics industry—where limited model capability and training data remain major hurdles.
Paris-based exoskeleton developer Wandercraft has unveiled its first humanoid robot, named Calvin, reportedly developed in just 40 days. Incorporating core components from Wandercraft’s proven exoskeleton technology, Calvin is a voice-controlled, self-balancing humanoid designed to handle physically demanding and ergonomically challenging industrial tasks. The company’s robust platform, originally tailored for medical use, is now being extended to manufacturing applications. This launch coincides with a strategic investment by Renault Group, which has entered into a commercial partnership with Wandercraft to advance the Calvin robot family. The collaboration aims to scale low-cost, high-impact robotics that enhance human life—whether by assisting individuals with mobility impairments or by supporting industrial automation in factories, clinics, and home environments.
Researchers from the federally funded Argonne National Laboratory in Illinois, in collaboration with the University of Virginia, have introduced an innovative technique for detecting defects in metal components fabricated through 3D printing. By integrating artificial intelligence, thermal imaging, and X-ray imaging, the team’s approach offers the potential for future real-time, self-correcting 3D printing systems. Specifically applied to laser powder bed fusion—a method that constructs metal parts layer by layer using a laser to melt metal powder—the system uses machine learning models to predict the formation of keyhole pores, a common defect in the process.
China has commenced mass production of the world’s first non-binary chips, marking a significant technological milestone with implications for sectors such as aviation and advanced manufacturing. Led by Professor Li Hongge and his research team at Beihang University in Beijing, the initiative integrates traditional binary logic with probabilistic logic, resolving key limitations in conventional computing architectures. This hybrid approach enables greater fault tolerance and improved power efficiency. The team is also designing a dedicated instruction set architecture (ISA) and microarchitecture optimized for hybrid probabilistic computing, which will enhance support for advanced workloads including AI acceleration, image and speech recognition, and neural network processing.
🧬 BioTech
Researchers in Australia have achieved a significant advancement in the quest to cure HIV, demonstrating the use of mRNA technology to expose dormant virus hidden within white blood cells. In a study published in Nature Communications, a Melbourne-based team showed for the first time that mRNA can be delivered directly into HIV-harboring cells using specially engineered lipid nanoparticles. Once inside, the mRNA reprograms the cells to reveal the virus, making it visible to the immune system or other therapies—an essential step toward fully eradicating HIV from the body.
Researchers at the University of Virginia have developed an interactive artificial pancreas system that leverages digital twin technology to manage type 1 diabetes. Early clinical trials have shown encouraging results. Known as Adaptive Biobehavioral Control (ABC), the system improves blood glucose regulation by dynamically adjusting to the user’s evolving physiological needs. According to Boris Kovatchev, PhD, director of the University’s Center for Diabetes Technology and lead author of the study, the system recalibrates the automated insulin delivery mechanism every two weeks for optimal performance. A key feature is a cloud-based digital twin that allows users to simulate various management strategies safely, providing a personalized environment to test potential changes before implementing them in real life. This represents the first application of digital twin modeling fully personalized to individual users in diabetes care.
A research team from the Chinese Academy of Sciences and the China Electric Power Research Institute has developed an advanced microrobot capable of precisely manipulating liquid droplets using magnetic fields. The microrobot can direct the movement of droplets, initiate controlled fusion for chemical reactions at low speeds, and perform high-speed splitting operations. Leveraging strong magnetic forces, the robot demonstrates speeds 20 times greater than its predecessors and can handle droplets nearly one milliliter in volume. The researchers emphasized their goal of creating a high-performance, residue-free system—an essential factor for applications in medical diagnostics and the handling of sensitive or reactive substances.
💡Products/tools of the week
Chat4Data is a Chrome extension designed to streamline web scraping using a conversational AI interface. It empowers users to extract structured information from websites without requiring any programming skills. By simply describing the type of data they need—such as product specifications, contact details, or multimedia—the tool intelligently identifies, collects, and formats the data for direct download as an Excel file. This efficient solution is especially valuable for professionals in market research, e-commerce, or academic fields who need to gather extensive web data without technical expertise.
Launched in May 2025, Skywork AI is redefining how professionals create content with its powerful, AI-driven workspace. Featuring a suite of specialized Super Agents for everything from documents and presentations to spreadsheets, web pages, and even podcasts, Skywork empowers users to produce research-rich content with ease. At the heart of the platform is its proprietary DeepResearch engine, delivering research up to 10x deeper than the competition and backed by verifiable sources. Ranked #1 on the GAIA benchmark, Skywork is the go-to solution for those who demand both speed and substance in creating high-quality, multimodal content.
PureCode AI is an enterprise-grade coding assistant designed to accelerate software development by automating both front-end and back-end workflows. Operating as a full-stack AI agent, it intelligently interprets your entire codebase to produce production-ready code, resolve bugs, modernize legacy systems, and generate UI components from simple text instructions. Development teams turn to PureCode AI to deliver features up to five times faster, minimize technical debt, and redirect their focus toward innovation rather than repetitive tasks. With seamless integration into VS Code and native support for popular frameworks such as React and Tailwind CSS, PureCode AI enhances productivity across the development lifecycle.
Tapflow is a purpose-built platform that empowers tech professionals and subject-matter experts to convert their knowledge into monetizable digital products such as guides, workflows, and courses. Designed for ease and efficiency, Tapflow simplifies the creation process so users can build and sell content without disrupting their daily work or needing to learn complex new tools. Professionals from fields like design, development, marketing, and education leverage Tapflow to generate new income streams, validate demand for their expertise, and share their knowledge in a polished, structured format. The platform includes features such as real-time collaboration, hierarchical content organization, and integrations for multiple payment systems to support seamless monetization.
