News Archive - Page 18 of 69

AI-generated science summaries promote understanding andn trust

In a world where science plays a big role in our everyday decisions, understanding it can be tricky. Scientists often write in a way that’s hard for most people to grasp, using lots of technical terms or jargon. This can make the public feel left out or distrustful of science because they can’t understand what’s being said. But a new study suggests a way to bridge this gap using artificial intelligence (AI).

David Markowitz, a professor at Michigan State University, has been looking into whether AI can make scientific writing easier to understand. His research, published in PNAS Nexus, shows that AI can indeed simplify complex scientific papers.

AI does this by creating summaries or “significance statements” that use simpler words and sentences. For example, instead of “occupation,” AI might just say “job,” which is easier for everyone to understand.

In his experiments, Markowitz found that when people read these AI-generated summaries, they understood the science better. They could explain what they read more accurately than those who only read the summaries written by scientists. This boost in comprehension also seemed to make readers trust and think more highly of the scientists involved.

The importance of better science communication

However, there’s a twist: when asked, people guessed that the simpler summaries were written by humans, not AI, showing a common misconception about how AI communicates.

While AI can make science more accessible, there are still things we need to learn. For one, as AI gets better, it could become a common tool in scientific journals, but we’re not there yet. There’s also the worry that simplifying too much might lose important details or introduce mistakes if not monitored. Plus, it’s important for readers to know when AI has been used to avoid any confusion or bias.

Ultimately, while AI offers a promising way to make science more digestible, scientists themselves could also improve how they communicate by using less jargon and clearer language, which might mean even without AI, we can all understand science better.

Markowitz summarized his findings in an article published in The Conversation.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

Ancient water on Mars

Scientists at Curtin University have found what might be the oldest signs of hot water activity on Mars. They looked at a very old rock from Mars, known as Black Beauty, which is a meteorite that landed on Earth. This rock, also called NWA7034, is about 4.45 billion years old and contains small crystals called zircon grains.

Zircon is a mineral that can survive intense conditions like heat and shock from meteor impacts. The scientists used nano-scale geochemistry to see and study very tiny parts of the zircon to find out what elements are in it.

What they found were traces of elements like iron, aluminium, yttrium, and sodium in the zircon. These elements are signs that water was around when the zircon formed. Water can change the composition of rocks, leaving behind these ‘fingerprints’ or chemical signs of its presence.

This process happens in hydrothermal systems, which are places where hot water from deep in the earth mixes with the rocks. On Earth, these systems are crucial for life because they provide warmth and nutrients.

This discovery tells us that Mars might have had the right conditions for life very early in its history, before 4.1 billion years ago. Even though Mars has gone through many big impacts from space rocks, this study shows that there was water during these ancient times, suggesting that Mars could have been habitable.

The study is published in Science Advances.

Chinese mission data suggests similar conclusions

In related research, data collected by the Chinese Mars rover Zhurong seems to confirm previous suggestions that an ocean may have existed in the northern lowland on Mars. The study, published in Nature, concludes that the Utopia Planitia region on Mars exhibits water-related features on its surface.

“In situ measurements by sensors onboard the Zhurong rover hardly provide direct evidence of the existence of an ancient ocean;” note the Chinese scientists. However, they do not contradict this model and could be considered within the context of ocean theory.”

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

First double lung transplant performed by robot

Doctors at NYU Langone Health have performed the world’s first double lung transplant using robots. They used machines to do the whole operation instead of doing it by hand. This is a big step forward because the procedure was less invasive.

The surgery was led by Stephanie Chang. She used a special robot called the da Vinci Xi to take out the old lungs and put in new ones for a 57-year-old woman named Cheryl Mehrkar. Cheryl had a lung disease called COPD, which makes it hard to breathe. Instead of the usual big cut, they made small cuts between her ribs. This way, the surgery is easier on the patient’s body.

Cheryl got her new lungs just four days after she was put on the list for a transplant. She had been sick for a long time, even more so after she had COVID-19.

Advantages of robotic surgery

The da Vinci Xi is “the most widely used multiport robotic surgery system in the world,” says the manufacturer. It “delivers proven capabilities for a wide spectrum of procedures across multiple specialties… offers advanced instrumentation, vision, and features such as Firefly fluorescence imaging and integrated table motion; it’s versatile and flexible, with standardization that can help manage inventory and improve overall OR efficiency.”

Chang explained that using robots for this kind of surgery can help patients feel less pain after the operation and recover faster. This is because the surgery is gentler on the body. “By using these robotic systems, we aim to reduce the impact this major surgery has on patients, limit their postoperative pain, and give them the best possible outcome,” says Chang in a NYU Langone Health press release.

“This latest breakthrough in robotic surgery speaks to the culture of innovation we’ve built by bringing the most talented people in their fields together,” adds Robert Montgomery, director of the NYU Langone Transplant Institute.

NYU Langone is now leading the way in this new type of lung surgery. They’ve already done a single lung transplant with robots before this.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

DeepSeek R1 reportedly thinks and reasons like OpenAI o1

A Chinese Artificial Intelligence (AI) lab named DeepSeek has introduced a large language model (LLM) called DeepSeek-R1, TechCrunch reports. According to DeepSeek, DeepSeek-R1 can think and reason much like OpenAI’s o1 model.

Reasoning means that DeepSeek-R1 doesn’t just rush to give quick answers. Instead, it takes its time to consider each question carefully, trying to avoid mistakes by essentially fact-checking itself. This involves thinking through tasks step by step, which often takes some time, especially for complex questions.

DeepSeek-R1 has been shown to perform as well as OpenAI’s o1 on two important tests: AIME, where AI models judge other AI models, and MATH, which involves solving word problems.

However, DeepSeek-R1 has certain flaws. For example, it has trouble with simple games like tic-tac-toe, a problem also seen with o1. Additionally, there are concerns about its security; people have found ways to make it bypass its safety features, like when someone got it to describe how to make methamphetamine.

Moreover, DeepSeek-R1 avoids answering questions that might be sensitive or controversial in China, like topics related to political figures or historical events, showing that it’s designed with certain restrictions to comply with local regulations.

Parts of the DeepSeek code are on Github.

Origins in automated stock trading

A Financial Times story (unpaywalled copy) published earlier this year revealed that the developer of DeepSeek is High-Flyer Capital Management, a Chinese quantitative hedge fund that is a big player in China’s financial sector.

The fund started exploring the potential of AI for automated stock trading. “AI helps to extract valuable data from massive data sets which can be useful for predicting stock prices and making investment decisions,” said one of the fund’s managers as reported by Financial Times.

The fund is one of six Chinese groups with more than 10,000 Nvidia A100 processors, often believed to be a computational threshold for self-training large models.

“High-Flyer aims to achieve ‘superintelligent’ AI,” TechCrunch reports without mentioning the source of this information.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

A dynamic crystal of free electrons in graphene

Electrons behave in unusual ways in pentalayer graphene, which is five layers of graphene stacked with boron nitride in a specific configuration. Graphene is a super-thin material made up of carbon atoms arranged in a hexagonal pattern.

Normally, when we think of electrons, we imagine them as whole units carrying a single negative charge. However, in some rare situations, electrons can split into smaller, fractional charges.

Typically, this splitting requires a very strong magnetic field. But electrons in pentalayer graphene showed this behavior even without a magnetic field.

The splitting is called the “fractional quantum anomalous Hall effect,” where “anomalous” means it happens without a magnetic field. MIT scientists have developed a new theory to explain this. They figured out that in this special setup, electrons aren’t just moving around freely; instead, they interact with each other due to being so closely packed in two-dimensional space.

This interaction leads them to form what could be described as a “crystal” of electrons, where their positions and movements become highly correlated.

This electron crystal isn’t like a normal crystal where atoms sit in fixed positions; it’s more like a dynamic cloud where electrons’ quantum states (or wavefunctions) are intertwined in a way that allows them to exhibit fractional charges.

Research directions for new quantum technologies

The MIT scientists have described the methods and results of this study in a paper published in Physical Review Letters.

“This crystal has a whole set of unusual properties that are different from ordinary crystals, and leads to many fascinating questions for future research,” says research leader Senthil Todadri in an MIT press release. “For the short term, this mechanism provides the theoretical foundation for understanding the observations of fractions of electrons in pentalayer graphene and for predicting other systems with similar physics.”

Understanding strange electron behavior in exotic materials could lead to new quantum technologies, where manipulation of quantum states at an atomic or molecular level is crucial.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

Living mouse created from gene older than animal life

Scientists at Queen Mary University and University of Hong Kong have created mouse stem cells from genes found in choanoflagellates.

Choanoflagellates are single-celled organisms that are distant ancestors of mice (and humans). The gene used in this study is called Sox.

Stem cells are special cells that can turn into any type of cell in the body, like skin, nerve, or muscle cells.

Choanoflagellates have genes similar to the Sox gene found in animals. Normally, these genes were thought to belong only to animals, but this research shows they existed in these simple organisms too.

The project “sounds like science fiction,” notes a Queen Mary University press release.

“By successfully creating a mouse using molecular tools derived from our single-celled relatives, we’re witnessing an extraordinary continuity of function across nearly a billion years of evolution,” says researcher Alex de Mendoza in the press release.

The scientists describe the methods and results of this study in a paper published in Nature Communications.

The scientists took the Sox gene from choanoflagellates and put it into mouse cells. This changed the mouse cells into stem cells that could grow into a mouse when put into a mouse embryo. So the ancient gene became part of a living mouse.

The mouse on the left is a chimeric with dark eyes and patches of black fur, a result of stem cells derived from a choanoflagellate Sox gene (*Credit: Gao Ya and Alvin Kin Shing Lee/CCMR*).

This mouse was special because it had traits from both the original mouse and the new stem cells, like having spots of black fur and dark eyes, proving that the ancient gene worked in animal development.

Potential applications to medicine and life sciences

This experiment suggests that the ability to make stem cells might have started very long ago, even before animals appeared on Earth. It shows how evolution recycles old genes for new purposes.

For example, while choanoflagellates don’t have stem cells, they use these genes for other basic life processes, which animals later adapted for their own complex needs.

This research isn’t just about history; it could help with future medical treatments. Understanding how these genes work could lead to better ways to use stem cells to treat diseases or heal injuries in humans.

This study suggests that synthetic versions of genes might perform even better than native animal genes in certain contexts, notes researcher Ralf Jauch.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

The sixth test flight of Starship was a major step toward operationality

SpaceX conducted the sixth test flight of Starship, aiming to push the boundaries of the rocket’s and booster’s capabilities towards making the entire system reusable.

The Super Heavy booster, equipped with all 33 Raptor engines, successfully lifted off from the pad.

Following a standard ascent, the booster separated from Starship as planned.

Since automated safety checks on the launch and catch tower detected an issue, SpaceX decided not to try and repeat the booster catch that was so spectacularly successful in the fifth text flight. The booster, following a pre-set safety protocol, safely diverted and performed a soft splashdown in the Gulf of Mexico.

Meanwhile, Starship continued its ascent, entering the expected trajectory. An important aspect of this test was the successful reignition of a single Raptor engine in space, a critical step for future missions where the spacecraft will need to deorbit.

With the aid of Starlink for live views and telemetry data, Starship navigated through reentry, executed a flip maneuver, performed a landing burn, and achieved a soft splashdown in the Indian Ocean.

The mission provided valuable data from several thermal protection experiments and the behavior of the spacecraft at high speeds and angles during reentry. These data are crucial for improving the design and operation of Starship, moving closer to the goal of routine and rapid reusability.

To the Moon and Mars

Despite not achieving the booster catch, the flight test was successful because it produced a lot of useful data that will be used for technical improvements, advancing SpaceX’s objectives for human and cargo transport to destinations like the Moon and Mars.

NASA Administrator Bill Nelson congratulated SpaceX, emphasizing the importance of the successful Raptor engine restart in space, which is a major progress towards orbital flight.

“Starship’s success is Artemis’ success,” said Nelson. “Together, we will return humanity to the Moon & set our sights on Mars.”

SpaceX hinted at a Starship test flight to Mars in 2026.

U.S. President elect Donald Trump went to Starbase to watch the launch.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

Faster hurricane modeling with machine learning

Hurricanes are powerful storms that can cause widespread destruction. Predicting their behavior, like how strong they’ll be or where they’ll hit land, is tough because these storms are influenced by many factors.

Researchers from City University of Hong Kong have used machine learning to better understand and predict hurricanes.

In their study, published in Physics of Fluids, the researchers focused on the boundary layer of the atmosphere. This is the part of the air closest to the Earth’s surface.

This layer is tricky to model because it interacts with the land, sea, and everything on the ground, making weather predictions challenging.

Traditional weather models use huge supercomputers and lots of data, but they often miss the mark.

“Our model employs an advanced physics-informed machine learning framework,” says researcher Feng Hu in a press release issued by American Institute of Physics.

He adds that the model requires only a small amount of real data to capture the complex behavior of the wind field of tropical cyclones. The model’s flexibility and ability to integrate sparse observational data result in more accurate and realistic reconstructions, he says.

Faster information for civil protection

The new machine learning method developed by the researchers uses equations from atmospheric physics to predict the wind patterns inside hurricanes. This method needs less data and can run faster than traditional models.

Understanding the wind field of a hurricane is crucial. It tells us about the storm’s power, its shape, and what kind of trouble it might bring to the shorelines. This model can provide a clearer view of what the hurricane’s wind will do, allowing to prepare better for its arrival.

The researchers believe that with climate change causing more frequent and intense hurricanes, their model could play an important role. This would mean better warnings for people in coastal areas, reducing damage and saving lives. They plan to expand this model to study other types of storms and integrate it into systems that provide real-time weather updates.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

AI finds cross-disciplinary connections and suggests research ideas

MIT scientists have developed a way for Artificial Intelligence (AI) to find connections between different scientific topics and suggest research ideas.

AI reproduces the results of human thought processes to a certain degree, but it usually struggles to link complex ideas from unrelated fields.

Generative AI can create new content, like text or music, by learning from existing data. The MIT scientists combined generative AI with graphs that show how things connect or relate to each other.

These graphs are based on category theory, a branch of abstract mathematics that models systems as collections of abstract objects and their relationships.

An open access paper published in Machine Learning: Science and Technology describes how the scientists made AI understand and reason about complex systems. The scientists trained AI systems with information from 1,000 scientific papers about biological materials, creating a map of knowledge.

An MIT press release summarizes some interesting connections found by the AI. An even more interesting one, analyzed in the paper, is found between flowers and nacre-inspired cement.

Microscopic view of nacre (Credit: Wikimedia Commons). — Microscopic view of nacre (*Credit: Wikimedia Commons*).

Nacre is an organic–inorganic composite material found in the inner shell of some mollusks. Nacre-inspired cement replicates some interesting properties of nacre in construction materials.

Creative research ideas

The scientists asked different large language models (LLM) to analyze multiple alternative paths between “a flower” and “nacre-inspired cement” in the knowledge graph. Then they asked the LLMs suggest creative research ideas.

The LLMs suggested creative research ideas indeed (see text boxes 1-4), with ChatGPT-4 showing “the most impressive reasoning capability and the most detailed response.”

All LLMs suggested to study the behavior of hydrogen and covalent bonds in certain materials. For examole, in chitosan (a natural polymer found in the exoskeleton of crustaceans) or polyethylene glycol dimethacrylate (PEGDMA).

GPT-4 also suggested to study how superhydrophobic surfaces, like those of rose petals, influence mechanical properties of materials.

This work shows that AI can help us think creatively across different fields, potentially leading to new discoveries.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

Robot wolves for military operations unveiled at Airshow China

At at 15th China International Aviation and Aerospace Exhibition, also known as Airshow China, Chinese developers unveiled an array of new military technology, focusing particularly on light arms and robotics, Global Times reports.

An interesting highlight was the presentation of a four-legged, wolf-like robot designed for military use, particularly in anti-terrorism roles.

A CCTV video posted to YouTube shows a pack of robot wolves in action.

These robots mimic the movement and teamwork of a wolf pack. They can navigate tough terrains, maintain stealth, and shoot with high precision, even at moving targets. They are equipped to carry various weapons like rifles or sub-machine guns, adapting to different mission requirements such as urban combat, border surveillance, or sniper operations.

The robot wolves operates in coordination with human soldiers, reducing the risk to human life during dangerous operations.

The robot’s actions can be monitored and controlled via screens, ensuring that in scenarios like hostage situations, it targets only threats, avoiding harm to innocent bystanders.

This technology represents a shift from individual robotic units to coordinated “packs,” echoing the cooperative and tactical nature of wolves in nature, thereby enhancing the effectiveness and safety of military operations.

China claims superiority in military robotics

There are multiple types of bionic robo-wolves “including integrated strike quadrupedal robots, reconnaissance and detection quadrupedal robots, transport quadrupedal robots, and operation and disposal quadrupedal robots,” Global Times reports in another article.

Britain, the US and other western armed forces are already testing and using robot dogs that can operate in a battlefield, The Times reports (unpaywalled copy). “But the Chinese army claims that its new pack of wolves is superior.”

“Compared to robodog, the robowolf has improved combat capabilities across various aspects, including reconnaissance, strike capability, and logistics support,” said Dai Jian of China South Industries Group Corporation, as reported by The Times.

“Its overall performance has been enhanced, which is why we call it a robowolf. Compared to a dog, it has stronger combat capabilities,” he added.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.

Exciting News! Our Mobile App is Here!

Welcome Back

No account? Create One

Join

Already have an account? Sign in

forgot password

The importance of better science communication

Chinese mission data suggests similar conclusions

Advantages of robotic surgery

Origins in automated stock trading

Research directions for new quantum technologies

Potential applications to medicine and life sciences

To the Moon and Mars

Faster information for civil protection

Creative research ideas

China claims superiority in military robotics