Weekly Piece of Future #92
From DIY Cancer Treatments to Autonomous Surgery and Human-Like Androids
Hey there, fellow future-addicts!
Welcome to this week's edition of Rushing Robotics, your go-to newsletter for the latest breakthroughs in artificial intelligence, biotechnology, and robotics. Here’s what’s on the horizon this week:
🤯 Mind-Blowing
In our opening segment, we explore some of the boldest moves in science and technology. A Croatian scientist's unconventional self-treatment for breast cancer, using lab-grown viruses instead of chemotherapy, raises profound questions about self-experimentation in medicine. We also highlight a new wave of robots trained solely by watching skilled professionals in action, such as surgical procedures—an approach pushing the boundaries of machine learning and imitation.
🔊 Industry Insights & Updates
Across the tech landscape, the largest companies are intensifying their focus on robotics and AI-driven innovation. Xpeng’s humanoid robot, Iron, stands as a remarkable achievement, blending advanced robotics with real-world manufacturing support. Meanwhile, a newly developed transparent semiconductor from the University of Minnesota promises to transform high-power electronics by combining efficiency with transparency, showing how AI and robotics are shaping industry applications in bold new directions.
🧬 BioTech
The field of biotechnology continues to reveal extraordinary advancements, from therapies that restore vision to diagnostic tools with the potential to anticipate viral outbreaks. A breakthrough stem cell therapy has already restored sight to several patients, while new genomic testing methods, developed for rapid pandemic response, could reshape how we diagnose infections. Meanwhile, an artificial muscle material from South Korea demonstrates unprecedented versatility, combining the softness of skin with the strength to handle loads as heavy as a car—an innovation with significant implications for robotics and wearable tech.
💡 Products/Tools of the Week
This week’s featured tools streamline complex processes, empowering innovators across various fields. Webdraw Alpha transforms hand-drawn sketches into fully functional web applications, bridging the gap between creativity and coding. Diaflow simplifies AI integration for businesses, offering a no-code editor that supports diverse applications, from chatbots to digital assistants. For developers, EarlyAI speeds up the process of creating unit tests by automatically generating test cases, making it easier than ever to deliver high-quality code.
🎥 Video Section
In this week’s curated video section, you’ll find the latest demonstrations in robotics and AI research. Watch as Unitree Robotics showcases robots achieving human-like walking within two days of learning, and see Meta’s AI research advancing robotic touch perception. Clone Robotics debuts a bimanual android torso with lifelike movement, while Johns Hopkins University demonstrates a surgical robot trained via machine learning. Finally, catch Engineai’s SE01 robot performing natural walking tests—offering a glimpse into a world where robotics moves closer to humanlike capabilities.
Each advancement reminds us that the future is for everyone to imagine, share, and shape. Stay hungry, stay futurish!
🤯 Mind-Blowing
A scientist’s groundbreaking self-treatment for stage 3 breast cancer, using lab-grown viruses instead of chemotherapy, has sparked an ethical debate over self-experimentation in medical science. In 2020, at age 49, Beata Halassy—a virologist from the University of Zagreb—discovered that her breast cancer had returned near the site of a previous mastectomy. Unwilling to undergo more chemotherapy, Halassy explored oncolytic virotherapy (OVT), a method that uses lab-grown viruses to target and destroy cancer cells. The case report, published in Vaccines, reveals that Halassy successfully treated herself and has remained cancer-free for four years. Her self-experimentation places her within a controversial tradition among scientists willing to test new treatments on themselves, a decision that Halassy noted took courage to publish. OVT is an emerging approach that deploys viruses to not only attack cancer cells but also activate immune defenses.
Clone Robotics, a Polish robotics startup, embarked on its journey by crafting a robotic hand with artificial muscles and bones, engineered to move like a human hand. Recently, the company showcased its next big step—a humanoid torso that can replicate realistic human movements. The demo video shows a ghostly white bot with movements lifelike enough to make an appearance in a horror movie, but the engineering within is impressive. While companies like Tesla are building robots to jump, dance, and assist with household tasks, Clone Robotics is focused on a different path. Founded in 2021, they specialize in biomimetic robotics, aiming to mimic the strength and dexterity of the human body, one of the most challenging designs to replicate.
A robot trained purely by watching experienced surgeons on video has successfully replicated intricate surgical procedures, demonstrating skills comparable to those of seasoned human doctors. Researchers suggest that using imitation learning to train surgical robots brings the field closer to fully autonomous robotic surgery, where robots might someday operate without human assistance. At Johns Hopkins University, researchers found that this training approach reduces the need to program every individual movement for each procedure. Similarly, Stanford University researchers used imitation learning to teach a da Vinci Surgical System robot three key tasks—needle handling, tissue lifting, and suturing—and the robot executed each with surgeon-like proficiency. This model combines machine learning architecture similar to that behind ChatGPT with imitation learning, though it operates in "robot" language, using kinematic expressions rather than words.
Researchers at TU Wien in Vienna have brought the future of housecleaning robots a step closer with a self-learning robot designed to mimic human actions for tasks like cleaning washbasins. While seemingly simple, programming a robot to navigate the complex curves of a sink with a sponge is incredibly challenging. Instead of hardcoding each motion, the team’s solution involved training the robot through observation and tactile data collected from human teachers. This approach could extend well beyond household chores, with potential applications in industries like polishing, painting, sanding, and adhesive application. The team achieved this by developing a sensor-equipped sponge, which human “teachers” used repeatedly on a sink’s edges to provide the robot with a model to follow.
The first-ever full-size purely-electric running robot, named Tiangong and developed in China, is now open source, a move intended to foster additional innovation and ease the robot’s integration into everyday applications. Tiangong, built by the Beijing Humanoid Robot Innovation Center, can sustain a running speed of 3.73 mph (6 km/h) and boasts sophisticated visual perception and 3D vision sensors, along with a high-precision IMU to tackle complex motion challenges. The robot was unveiled on May 4 at the Beijing Economic-Technological Development Area, marking a milestone in robotics.
🔊 Industry Insights & Updates
Xpeng, the leading Chinese electric vehicle manufacturer, has unveiled its first humanoid robot, Iron, at its 2024 AI Day event. This marks a major expansion into the robotics field, much like Tesla's own efforts with humanoid robots. Iron stands 5'8" and weighs 154 pounds (70 kilograms), featuring over 60 joints and 200 degrees of freedom. The robot is already helping assemble the new P7+ electric vehicle on Xpeng’s automotive production lines, showcasing its practical applications in manufacturing. Iron's capabilities, powered by Xpeng's Turing AI chip, include walking and handling intricate tasks with precision, positioning the company as a key player in the convergence of AI and robotics.
A new transparent and efficient semiconductor material developed by researchers at the University of Minnesota could transform high-power electronics, particularly devices that require transparency, such as lasers. This fully synthetic material supports rapid electron movement while remaining transparent to visible and UV light. Already setting records, it’s considered a significant leap forward in semiconductor technology. As the trillion-dollar semiconductor industry expands, advances in ultra-wide band gap materials, which efficiently conduct electricity in challenging environments, are crucial for the evolution of digital tech used in phones, medical equipment, and more.
An AI agent has made a groundbreaking discovery of a previously unknown, zero-day memory-safety vulnerability in widely used real-world software, marking a first in the public release of such a finding. This significant achievement was made possible by the collaboration between Google’s Project Zero and DeepMind, who worked together to develop the Big Sleep large language model-assisted vulnerability agent. According to the Project Zero blog, this new agent was able to identify an exploitable stack buffer underflow in SQLite, a widely used open-source database engine. This marks the first time an AI-powered agent has successfully found a zero-day vulnerability in real-world code.
Researchers at the University of Pennsylvania are pioneering advancements in robotics by harnessing radio signals to grant robots a form of superhuman sight. Their cutting-edge technology, PanoRadar, transforms simple radio waves into detailed 3D visuals, enabling robots to "see" beyond conventional sensor boundaries. By processing radio waves with AI, this device enhances the low-quality imagery typical of standard radar. This breakthrough allows robots to accurately navigate through obstacles such as smoke, glass, and walls—barriers that often stymie traditional sensors. This AI-driven perception innovation could greatly improve multi-modal systems, supporting robots in complex tasks like search and rescue or autonomous driving.
🧬 BioTech
The promise of stem cell therapy for vision restoration has made a remarkable leap forward with the success of the world’s first vision-restoring stem cell treatment. Three individuals, each suffering from profound vision impairment, have seen dramatic improvements in their eyesight after receiving stem cell transplants. These improvements have lasted for over a year, representing a major breakthrough in the field of regenerative medicine and offering new hope for those who have lost their sight. A fourth person, with very compromised vision, also showed initial improvement, though the positive effects were not sustained. These four patients are the first to receive transplants of reprogrammed stem cells aimed at repairing and healing the cornea, which is the transparent, protective outer surface of the eye. This groundbreaking treatment could lead to future therapies that restore vision in people suffering from corneal damage and other causes of blindness, signaling a new era in eye care.
A revolutionary genomic test has been developed to diagnose nearly any infection, with far-reaching implications for diagnosing neurological infections such as meningitis and encephalitis, as well as quickly detecting new viral threats like pandemics. The test leverages a powerful sequencing method called metagenomic next-generation sequencing (mNGS), and the research team demonstrated that it correctly identified 86% of neurological infections. The test can also identify respiratory viruses with pandemic potential—including SARS-CoV-2, influenza A and B, and RSV—in less than a day, even with small amounts of virus in the sample. In addition, the team’s modeling showed that the test could detect new, divergent viral strains should they emerge in the future. Both versions of the mNGS test—one for cerebrospinal fluid (CSF) and one for respiratory samples—have received breakthrough device designation from the U.S. Food and Drug Administration.
A research team at South Korea's Ulsan National Institute of Science & Technology (UNIST) has engineered an innovative artificial muscle material that is as soft as human skin yet incredibly strong. This magnetic composite can adapt its stiffness, transitioning smoothly from soft to rigid and back. Remarkably, this artificial muscle can handle loads similar to those of a car and is 2,700 times stiffer than traditional materials. The development, which involves ferromagnetic particles and shape memory polymers, could be transformative for soft robotics and wearable tech. The ferromagnetic particles make it possible to control the muscle remotely while also adding to its strength.
💡Products/tools of the week
Webdraw Alpha offers a revolutionary approach to web application development by transforming hand-drawn sketches into executable code. Designed to eliminate technical barriers, this platform allows users to create functional web apps through a simple drawing interface, bypassing the need for in-depth coding expertise. It’s a valuable tool for founders, engineers, product managers, and creators who want to turn their ideas into reality with speed and ease. Webdraw Alpha streamlines the development process, making web app creation more intuitive, faster, and user-friendly, enabling users to focus on innovative design and performance.
Diaflow is an enterprise-focused platform that makes AI integration and deployment straightforward across multiple business areas. Its no-code editor allows users to create AI-powered applications, manage workflows, and organize data efficiently, even without coding skills. With Diaflow, organizations can develop chatbots, digital assistants, and internal tools, supporting departments like marketing, operations, and IT. Known for its ease of use, speed, and extensive integrations, Diaflow offers a scalable, cost-effective AI solution for enterprises eager to foster innovation and improve operational efficiency.
EzyGraphs is a convenient and accessible graphing app, available both on mobile devices and as a web app, designed for easy mathematical visualization. With a user-friendly interface, it accurately displays mathematical notations and graphs various functions, making it ideal for students, teachers, and professionals seeking quick graphing capabilities. EzyGraphs is a practical choice for anyone who wants simple yet effective graphing without the complexity of advanced software, offering portability and educational value for on-the-go math visualization.
EarlyAI leverages artificial intelligence to automatically generate high-quality unit tests for developers. It integrates into IDEs to create complete test suites, including mocks and edge cases, while enhancing code coverage. Developers benefit from EarlyAI’s ability to save time, improve code quality, and streamline Test-Driven Development (TDD). Whether you're an individual developer or part of a team or enterprise, EarlyAI simplifies testing processes and accelerates the delivery of reliable, quality software.
