Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 178

Utility Engineering

As AIs rapidly advance and become more agentic, the risk they pose is governed not only by their capabilities but increasingly by their propensities, including goals and values. Tracking the emergence of goals and values has proven a longstanding problem, and despite much interest over the years it remains unclear whether current AIs have meaningful values. We propose a solution to this problem, leveraging the framework of utility functions to study the internal coherence of AI preferences. Surprisingly, we find that independently-sampled preferences in current LLMs exhibit high degrees of structural coherence, and moreover that this emerges with scale. These findings suggest that value systems emerge in LLMs in a meaningful sense, a finding with broad implications. To study these emergent value systems, we propose utility engineering as a research agenda, comprising both the analysis and control of AI utilities. We uncover problematic and often shocking values in LLM assistants despite existing control measures. These include cases where AIs value themselves over humans and are anti-aligned with specific individuals. To constrain these emergent value systems, we propose methods of utility control. As a case study, we show how aligning utilities with a citizen assembly reduces political biases and generalizes to new scenarios. Whether we like it or not, value systems have already emerged in AIs, and much work remains to fully understand and control these emergent representations.

Scientists develop novel self-healing electronic skin for health monitoring

Researchers have achieved a breakthrough in wearable health technology by developing a novel self-healing electronic skin (E-Skin) that repairs itself in seconds after damage. This could potentially transform the landscape of personal health monitoring.

In a study published in Science Advances, scientists demonstrate an unprecedented advancement in E-Skin technology that recovers over 80% of its functionality within 10 seconds of being damaged—a dramatic improvement over existing technologies that can take minutes or hours to heal.

The technology seamlessly combines ultra-rapid self-healing capabilities, reliable performance in , advanced artificial intelligence integration, and highly accurate health monitoring systems. This integration enables real-time fatigue detection and muscle strength assessment with remarkable precision.

New device helps people with spinal cord injuries grasp large, heavy objects

More than 15 million people worldwide are living with spinal cord injury (SCI), which can affect their sensory and motor functions below the injury level. For individuals with SCI between C5 and C7 cervical levels, this can mean paralysis affecting their limbs and limited voluntary finger and wrist flexion, making it difficult to grasp large, heavy objects.

Now, a team of UC Berkeley engineers from the Embodied Dexterity Group has developed a to enhance grasping functionality in this population. Dubbed the Dorsal Grasper, this leverages voluntary wrist extension and uses supernumerary robotic fingers on the back of the hand to facilitate human-robot collaborative grasping.

In a study recently featured in IEEE Transactions on Neural Systems and Rehabilitation Engineering, the researchers demonstrated for the first time how the Dorsal Grasper can expand users’ graspable workspace. Test subjects found that they could easily grasp objects anywhere they could reach their arm, without having to rotate their bodies, which can cause wheelchair users to lose their balance.

AI program plays the long game to solve decades-old math problems

A game of chess requires its players to think several moves ahead, a skill that computer programs have mastered over the years. Back in 1996, an IBM supercomputer famously beat the then world chess champion Garry Kasparov. Later, in 2017, an artificial intelligence (AI) program developed by Google DeepMind, called AlphaZero, triumphed over the best computerized chess engines of the time after training itself to play the game in a matter of hours.

More recently, some mathematicians have begun to actively pursue the question of whether AI programs can also help in cracking some of the world’s toughest problems. But, whereas an average game of chess lasts about 30 to 40 moves, these research-level math problems require solutions that take a million or more steps, or moves.

In a paper appearing on the arXiv preprint server, a team led by Caltech’s Sergei Gukov, the John D. MacArthur Professor of Theoretical Physics and Mathematics, describes developing a new type of machine-learning algorithm that can solve math problems requiring extremely long sequences of steps. The team used their to solve families of problems related to an overarching decades-old math problem called the Andrews–Curtis conjecture. In essence, the algorithm can think farther ahead than even advanced programs like AlphaZero.

Dual-domain architecture shows almost 40 times higher energy efficiency for running neural networks

Combining concepts from statistical physics with machine learning, researchers at the University of Bayreuth have shown that highly accurate and efficient predictions can now be made as to whether a substance will be liquid or gaseous under given conditions. They have published their findings in Physical Review X.

Observation of a glass of water reveals that the water exists in two : liquid and gas. Even at room temperature, water molecules are constantly evaporating from the surface of the liquid water and passing into the gas phase. At the same time, some of the water molecules from the gas condense back into the liquid.

The transition from one phase to the other depends on temperature and pressure. Above a , the simultaneous coexistence of gas and liquid disappears. The resulting supercritical fluid no longer forms an interface. This is important for industrial processes such as separation, cleaning and production.

Untethered soft actuators for soft standalone robotics

Soft actuators produce the mechanical force needed for the functional movements of soft robots, but they suffer from critical drawbacks since previously reported soft actuators often rely on electrical wires or pneumatic tubes for the power supply, which would limit the potential usage of soft robots in various practical applications. In this article, we review the new types of untethered soft actuators that represent breakthroughs and discuss the future perspective of soft actuators. We discuss the functional materials and innovative strategies that gave rise to untethered soft actuators and deliver our perspective on challenges and opportunities for future-generation soft actuators.

For pneumatic actuators, the pneumatic pumps serve an essential role in generating a mechanical force by using compressed gas or moving the liquid for the rapid fluid pressure increase. Yet, the incorporation of the pneumatic pump into the soft robotics would impair the mobility and the core functionalities of the soft robots because the pumps are usually relatively bulky and heavy when compared to the soft robots themselves. To address this issue, several recent studies demonstrated pump-less pneumatic actuation by employing the phase change materials that generate the volume change as the materials switch between liquid and gaseous states, thus resulting in the inflation and deflation of actuators. Here, the pump-less pneumatic actuators can be defined as the soft actuators that do not use the actual pump but generate a pneumatic force by the phase change of material just as if utilizing the pneumatic pump. In other words, the pump-less pneumatic actuators just reproduce the end effect of the pump by a different working mechanism without using the actual pump. The absence of the pneumatic pump in the robotic design also eliminates the need for pneumatic tubes to infuse/extract air into/from the actuator, thereby making the design completely untethered.

Likewise, external stimuli can deliver a considerable amount of mechanical displacement and force needed to actuate the soft robots in an untethered manner: the external stimuli in this article include magnetic field, heat, electricity, light, and humidity. Hence, without physically connecting the electrical tethering to the soft actuators to provide the power source, the external stimuli can enable the soft actuator to produce mechanical displacement since the materials are designed to actuate as programmed. As opposed to the pneumatics-based soft actuators that require the onboard power source (such as a battery or self-powering energy harvesting devices) to supply power to induce pneumatic force, some of these actuators receive the power to induce mechanical displacement in a completely untethered fashion. For example, systematic manipulation of a magnetic field can control the movement of the magnet-driven soft actuator as intended without any type of wiring. Similarly, if the antennas are incorporated into the soft robotic system, electromagnetic waves can be utilized to provide power wirelessly to operate the soft actuators5,6,7,8, or it also enables the remote control of the actuators via wireless communication9,10 Therefore, external stimuli-driven soft actuators retain the potential to represent the breakthrough in the field of soft robotics although there exist considerable limitations to be resolved. In this light, it would be a highly valuable resource to introduce untethered soft actuators and discuss the future perspective of new types of soft robots. There are a considerable number of review articles on soft actuators and robotics11,12,13,14,15. However, no review paper has dealt with recent advances in untethered soft actuators for soft robotics that demonstrated meaningful outcomes within a few years. Recently, roboticists and researchers proposed an explosive number of soft actuators for soft robots based on innovative structural designs and functional materials that represent breakthroughs in the field of soft robotics. Furthermore, as the field of soft actuators is relatively new and drawing a substantial amount of interest in the related fields, there exists a demand for an article that systematically reviews the current trend and informs the opportunities to contribute to the field. In this regard, we believe that the timely and thorough review of the recent advances in untethered soft actuators will be informative for the general readers who wish to draw insights and gain potential perspectives in the field.

In this article, we introduce the representative works of the untethered soft actuators that serve as breakthroughs in soft robotics and further discuss the imminent challenges of the soft actuators to be addressed. Soft actuators can also be applied to rigid robots since the actuators reviewed in this paper operate in an untethered configuration. However, we intentionally circumscribed the scope and focused mainly on soft actuators for soft robots because the incorporation of soft actuators into the soft robot can make the entire robot soft and compliant. There exist specific applications where the soft robots exhibit comparative strengths over rigid robots such as navigating through the tortuous space16,17, exploring deep-sea at extremely high pressure18, or minimally invasive surgery19. Furthermore, to present these works systematically, the paper categorizes the soft actuators by four representative working mechanisms (1. pneumatically/hydraulically-driven, 2. magnetically-driven, 3. heat-driven, and 4. electrically-driven) and further examines each actuating mechanism in relation to the untethered soft robots as illustrated in Fig. 1. The paper examines the strengths and limitations of each actuating method and concludes with the future perspective of untethered soft actuators for soft robotics. Box 1 provides the general summary that addresses the strategies to provide the power source for actuation control of the soft robots. Additionally, Table 1 draws the overall comparison of each soft actuating method to highlight the strengths, weaknesses, and other key features such as response time and output force range. On the other hand, Table 2 captures key highlights of representative soft actuators that operate based on a variety of mechanisms and thus delivers a more specific comparison.

Scientists crack how cat-like tail could make robots move like ninjas

Rigid, lizard-like tails are simple, but mammal-style tails may be lighter and better for space. Researchers studied how tails aid midair maneuverability in animals and robots, focusing on inertial appendages that generate body rotation.

Inspired by lizards and geckos, roboticists have designed rigid, single-plane tails to enhance stability and control in aerial and terrestrial robots. Some robotic tails aid in landing, flight reorientation, and high-speed turns. However, vertebrate (like cats and squirrels) tails are more complex, consisting of multiple vertebrae that allow for diverse movements.

By analyzing mammalian tails, researchers found that increasing bone segments within the same length enhances rotational ability. To evaluate tail effectiveness, they developed simulations optimizing tail trajectories for precise body rotations. Unlike previous models that assumed rigid structures, their approach considers deformability and realistic control constraints.

Quantum Meets AI: New Research Uncovers Novel Photochromic Materials

A joint research team has developed an innovative quantum-classical computing approach to design photochromic materials—light-sensitive compounds—offering a powerful tool to accelerate material discovery. Their findings were published in Intelligent Computing.

Building on their previous work in the same journal, the researchers introduced a computational-basis variational quantum deflation method as the foundation of their approach.

To validate its effectiveness, the team conducted a case study in photopharmacology, screening 4,096 diarylethene derivatives. They identified five promising candidates that exhibited two critical properties: large maximum absorbance wavelengths and high oscillator strengths. These characteristics are crucial for applications such as light-controlled drug delivery in photopharmacology.