Even though virtual reality headsets are popular for gaming, they haven’t yet become the go-to device for watching television, shopping, or using software tools for design and modeling. One reason why is because VR can make users feel sick with nausea, imbalance, eye strain, and headaches. This happens because VR creates an illusion of 3D viewing — but the user is actually staring at a fixed-distance 2D display. The solution for better 3D visualization exists in a 60-year-old tech that’s being updated for the digital world — holograms.
A new method called tensor holography enables the creation of holograms for virtual reality, 3D printing, medical imaging, and more — and it can run on a smart-phone. Holograms deliver an exceptional 3D representation of the world around us — and they’re beautiful. Holograms offer a shifting visual perspective based on the viewer’s position. They allow the human eye to adjust its focal depth — so you can move your focus easily from foreground to background. The visual holographic display appears just like a real 3D object — as if you could touch it.
Researchers have been trying to make computer-generated holograms. But the process has traditionally required a super-computer to churn through heaps of physics simulations — to create that life-like effect. That’s time-consuming and can yield less-than-photo-realistic results.
Holograms in the blink of an eye.
To deal with this, researchers at the Massachusetts Institute of Technology (MIT) designed a new way to produce holograms — almost instantly. The software they’re using is called a deep learning artificial intelligence program, because it can teach itself. They said it’s so efficient that it can create a hologram on a laptop — in the blink of an eye.
Liang Shi is a PhD student at MIT and the lead researcher on the project. He said: "People previously thought that with existing consumer-grade hardware, it was impossible to do real-time 3D holography computations. It’s often said that commercially available holographic displays will be around in 10 years — but this statement has been around for decades. This new approach — tensor holography — will finally bring that elusive 10-year goal within reach. This advance could fuel a spill-over of holography into fields like VR and 3D printing."
NOTE: All articles in the amazing-science newsletter can also be sorted by topic. To do so, click the FIND button (symbolized by the FUNNEL on the top right of the screen) and display all the relevant postings SORTED by TOPICS.
Photonic components, including single photon sources, single-photon detectors, and an ultra-high-performance optical switch, are fabricated in a high-volume semiconductor foundry on standard silicon wafers. Photonic qubits avoid exotic materials and manufacturing processes and are able to leverage the existing reliability, volume and precision of standard semiconductor manufacturing processes.
Photonic qubits are implemented by repurposing integrated photonics technology, originally developed for telecom and datacenter networking applications. Entangled states — specially designed to implement quantum error-correcting codes — are created and measured using fusion gates. Nondeterministic photon sources and gate operations are made scalable via a combination of multiplexing and loss-tolerant error correcting codes.
Photonic chips are integrated with cryogenic control electronics into a manufacturable package. Despite the fact that some of these packages must be operated at cryogenic temperature, we are able to predominantly use conventional technologies — PCBs, copper heat sinks and bump bonding. As a result, PsiQuantum can leverage conventional contract manufacturing to build large numbers of packages quickly. Photonic qubits can be networked and modularized using standard optical fiber, without transduction — a huge advantage at scale.
Photonic qubits do not feel heat, do not decohere and can be operated at high fidelity at room temperature. While the superconducting detectors needed for high-efficiency readout of photonic qubits do require cryogenic cooling, they operate at ~4K, relative to ~10mK (e.g. superconducting qubits). PsiQuantum is building and testing cryogenic cabinets with thousands of times more cooling power than the largest dilution refrigerators ever built or operated. LINK Because of the higher operating temperature of photonic qubits, these cryostats can also be surprisingly simple and can be built using existing cryogenic technology.
Most quantum computing architectures will require a datacenter-like facility in order to deliver enough qubits for commercially useful applications. PsiQuantum has partnered with both Daresbury SFTC in the UK and the SLAC National Accelerator Laboratory in California to validate intermediate-scale systems. In April 2024, PsiQuantum announced it will build the first utlity-scale, fault-tolerant quantum computer in Brisbane, Australia as part of an investment and partnership with the Australian Commonwealth and Queensland governments.
The applied software enables programming quantum computers using a familiar language: Python. It affords programmers the ability to specify algorithms at a high level using contemporary primitives such as quantum signal processing, qubitization and block encoding, and automates the compilation to individual gate operations. PsiQuantum have a growing library of over 50 proven quantum algorithms and subroutines that are callable by any application, reducing development time and improving the quality of domain-specific applications. While PsiQuantum has a very fast qubit simulator, its main emphasis is optimization of resource requirements for fault-tolerant implementation of quantum algorithms. They are working with partners from industry, academia and government to develop and optimize quantum algorithms for a broad range of valuable applications.
NVIDIA founder and CEO Jensen Huang delivers keynote address at COMPUTEX 2024 on June 2, 2024 at 7 p.m. in Taipei, Taiwan, outlining what’s next for AI and NVIDIA.
Kling is a video generation model developed by the Kuaishou Big Model Team. It has powerful video generation capabilities, allowing users to easily and efficiently create artistic videos.
Speaker in Paris: Romain Huet, OpenAI, Head of Developer Experience
The Viva Technology or VivaTech conference is an annual startup and technology event held in Paris, France. It is the conference that accelerates innovation by connecting major corporations, tech leaders, startups, and investors to the world's biggest challenges.
VivaTech organizes the biggest startup and technology event in Europe every year, involving four electrifying days in Paris. It unites the rule-breaking entrepreneurs, tech-related disruptive topics, and the largest technology breakthroughs.
VivaTech, on its own, is a global community that involves thousands of investors, visionary startups, researchers, corporations, talent, media, and organizations. That's why the Viva Technology conference is the much-awaited and Europe's biggest technology and startup event.
The 2023 Viva Technology event was held in Paris from June 14 to 17. It involved over 150,000 attendees and thousands of exhibitors and startups. Many keynote speakers were present, including Emmanuel Macron (the President of the French Republic), Elon Musk (Head of SpaceX, Tesla, and X), Dan Schulman (president of PayPal), and many more.
NASA’s Innovative Advanced Concepts program (NIAC) has selected six visionary concept studies for additional funding and development. Each study has already completed the initial NIAC phase, showing their futuristic ideas – like a lunar railway system and fluid-based telescopes – may provide fresh perspectives and approaches as NASA explores the unknown in space.
The NIAC Phase II conceptual studies will receive up to $600,000 to continue working over the next two years to address key remaining technical and budget hurdles and pave their development path forward. When Phase II is complete, these studies could advance to the final NIAC phase, earning additional funding and development consideration toward becoming a future aerospace mission.
“These diverse, science fiction-like concepts represent a fantastic class of Phase II studies,” said John Nelson, NIAC program executive at NASA Headquarters in Washington. “Our NIAC fellows never cease to amaze and inspire, and this class definitely gives NASA a lot to think about in terms of what’s possible in the future.”
The six concepts chosen for 2024 NIAC Phase II awards are:
Fluidic Telescope (FLUTE): Enabling the Next Generation of Large Space Observatorieswould create a large optical observatory in space using fluidic shaping of ionic liquids. These in-space observatories could potentially help investigate NASA’s highest priority astrophysics targets, including Earth-like exoplanets, first-generation stars, and young galaxies. The FLUTE study is led by Edward Balaban from NASA’s Ames Research Center in California’s Silicon Valley.
The Great Observatory for Long Wavelengths (GO-LoW)could change the way NASA conducts astronomy. This mega constellation low-frequency radio telescope uses thousands of autonomous SmallSats capable of measuring the magnetic fields emitted from exoplanets and the cosmic dark ages. GO-LoW is led by Mary Knapp with MIT in Cambridge, Massachusetts.
Radioisotope Thermoradiative Cell Power Generatoris investigating new in-space power sources, potentially operating at higher efficiencies than NASA legacy power generators. This technology could enable small exploration and science spacecraft in the future that are unable to carry bulky solar or nuclear power systems. This power generation concept study is from Stephen Polly at the Rochester Institute of Technology in New York.
FLOAT: Flexible Levitation on a Trackwould be a lunar railway system, providing reliable, autonomous, and efficient payload transport on the Moon. This rail system could support daily operations of a sustainable lunar base as soon as the 2030s. Ethan Schaler leads FLOAT at NASA’s Jet Propulsion Laboratory in Southern California.
ScienceCraft for Outer Planet Explorationdistributes Quantum Dot-based sensors throughout the surface of a solar sail, enabling it to become an innovative imager. Quantum physics would allow NASA to take scientific measurements through studying how the dots absorb light. By leveraging the solar sail’s area, it allows lighter, more cost-effective spacecraft to carry imagers across the solar system. ScienceCraft is led by NASA’s Mahmooda Sultana at the agency’s Goddard Space Flight Center in Greenbelt, Maryland.
NASA’s Space Technology Mission Directorate funds the NIAC program, as it is responsible for developing the agency’s new cross-cutting technologies and capabilities to achieve its current and future missions.
Machine-learning system trained on millions of human audio clips shows promise for detecting COVID-19 and tuberculosis. A team led by Google scientists has developed a machine-learning tool that can detect and monitor health conditions by evaluating noises such as coughing and breathing. The artificial intelligence (AI) system, trained on millions of audio clips of human sounds, might one day be used by physicians to diagnose diseases including COVID-19 and tuberculosis and to assess how well a person’s lungs are functioning. This is not the first time a research group has explored using sound as a biomarker for disease. The concept gained traction during the COVID-19 pandemic, when scientists discovered that it was possible to detect the respiratory disease through a person’s cough.
What’s new about the Google system — called Health Acoustic Representations (HeAR) — is the massive data set that it was trained on, and the fact that it can be fine-tuned to perform multiple tasks. The researchers, who reported the tool earlier this month in a preprint1 that has not yet been peer reviewed, say it’s too early to tell whether HeAR will become a commercial product. For now, the plan is to give interested researchers access to the model so that they can use it in their own investigations. “Our goal as part of Google Research is to spur innovation in this nascent field,” says Sujay Kakarmath, a product manager at Google in New York City who worked on the project....
The COVID-19 pandemic has affected hundreds of millions of individuals and caused more than six million deaths. The prolonged pandemic duration and the continual inter-individual transmissibility have contributed to the emergence of a wide variety of SARS-CoV-2 variants. Genomic surveillance and phylogenetic studies have shown that substantial mutations in crucial supersites of spike glycoprotein modulate the binding affinity of the evolved SARS-COV-2 lineages to ACE2 receptors and modify the binding of spike protein with neutralizing antibodies.
The immunological spike mutations have been associated with differential transmissibility, infectivity, and therapeutic efficacy of the vaccines and the immunological therapies among the new variants. This review highlights the diverse genetic mutations assimilated in various SARS-CoV-2 variants. The implications of the acquired mutations related to viral transmission, infectivity, and COVID-19 severity are discussed. This review also addresses the effectiveness of human neutralizing antibodies induced by SARS-CoV-2 infection or immunization and the therapeutic antibodies against the ascended variants.
Using a virus-like delivery particle made from DNA, researchers from MIT and the Ragon Institute of MGH, MIT, and Harvard have created a vaccine that can induce a strong antibody response against SARS-CoV-2.
The vaccine, which has been tested in mice, consists of a DNA scaffold that carries many copies of a viral antigen. This type of vaccine, known as a particulate vaccine, mimics the structure of a virus. Most previous work on particulate vaccines has relied on protein scaffolds, but the proteins used in those vaccines tend to generate an unnecessary immune response that can distract the immune system from the target.
In the mouse study, the researchers found that the DNA scaffold does not induce an immune response, allowing the immune system to focus its antibody response on the target antigen. “DNA, we found in this work, does not elicit antibodies that may distract away from the protein of interest,” says Mark Bathe, an MIT professor of biological engineering. “What you can imagine is that your B cells and immune system are being fully trained by that target antigen, and that’s what you want — for your immune system to be laser-focused on the antigen of interest.”
Viral infection can lead to acute respiratory distress syndrome that damages the lungs and is often lethal. Zhao et al. show that the TGF-β receptor 2 (TGF-βR2) is important for lung endothelial cells to recover after injury due to influenza virus or SARS-CoV-2 infection. The researchers used mice, human organoids, and human endothelial cells to demonstrate how TGF-βR2 interacts with VEGFA specifically in endothelial lung cells to promote tissue regeneration after viral injury. They also developed lung-targeted nanoparticles to deliver Vegfa mRNA to promote lung healing in mice lacking TGF-βR2. This study highlights the potential for targeting TGF-βR2 signaling in lung endothelial cells as a treatment for acute respiratory distress syndrome.
Disruption of pulmonary vascular homeostasis is a central feature of viral pneumonia, wherein endothelial cell (EC) death and subsequent angiogenic responses are critical determinants of the outcome of severe lung injury. A more granular understanding of the fundamental mechanisms driving reconstitution of lung endothelium is necessary to facilitate therapeutic vascular repair.
Now a team of scientists demonstrated that TGF-β signaling through TGF-βR2 (transforming growth factor–β receptor 2) is activated in pulmonary ECs upon influenza infection, and mice deficient in endothelial Tgfbr2 exhibited prolonged injury and diminished vascular repair. Loss of endothelial Tgfbr2 prevented autocrine Vegfa (vascular endothelial growth factor α) expression, reduced endothelial proliferation, and impaired renewal of aerocytes thought to be critical for alveolar gas exchange.
Angiogenic responses through TGF-βR2 were attributable to leucine-rich α-2-glycoprotein 1, a proangiogenic factor that counterbalances canonical angiostatic TGF-β signaling. Further, the scientists developed a lipid nanoparticle that targets the pulmonary endothelium, Lung-LNP (LuLNP). Delivery of Vegfa mRNA, a critical TGF-βR2 downstream effector, by LuLNPs improved the impaired regeneration phenotype of EC Tgfbr2 deficiency during influenza injury.
These studies defined a role for TGF-βR2 in lung endothelial repair and demonstrated efficacy of an efficient and safe endothelial-targeted LNP capable of delivering therapeutic mRNA cargo for vascular repair in influenza infection.
Armed with a catalog of hundreds of thousands of DNA and RNA virus species in the world's oceans, scientists are now zeroing in on the viruses most likely to combat climate change by helping trap carbon dioxide in seawater or, using similar techniques, different viruses that may prevent methane's escape from thawing Arctic soil.
By combining genomic sequencing data with artificial intelligence analysis, researchers have identified ocean-based viruses and assessed their genomes to find that they "steal" genes from other microbes or cells that process carbon in the sea. Mapping microbial metabolism genes, including those for underwater carbon metabolism, revealed 340 known metabolic pathways throughout the global oceans. Of these, 128 were also found in the genomes of ocean viruses.
"I was shocked that the number was that high," said Matthew Sullivan, professor of microbiology and director of the Center of Microbiome Science at The Ohio State University. Having mined this massive trove of data via advances in computation, the team has now revealed which viruses have a role in carbon metabolism and are using this information in newly developed community metabolic models to help predict how using viruses to engineer the ocean microbiome toward better carbon capture would look.
"The modeling is about how viruses may dial up or dial down microbial activity in the system," Sullivan said. "Community metabolic modeling is telling me the dream data point: which viruses are targeting the most important metabolic pathways, and that matters because it means they're good levers to pull on."
If fully successful, the IM-1 cargo mission would be the first U.S. lunar landing in more than 50 years.
Intuitive Machines’ Nova-C lunar lander launched from Florida on SpaceX’s Falcon 9 rocket, beginning the IM-1 mission.
If fully successful, the IM-1 cargo mission would be the first U.S. lunar landing in more than 50 years.
The Intuitive Machines lander is expected to spend about eight days traveling to the moon before descending to the surface.
Intuitive Machines, Inc. is a diversified space company focused on space exploration. It is a provider and supplier of space products and services that enable sustained robotic and human exploration to the Moon, Mars, and beyond. Its products and services are offered through its four business units: Lunar Access Services, Orbital Services, Lunar Data Services and Space Products and Infrastructure. Its Orbital Services segment is designed to support satellites and stations in Earth and lunar orbits. Orbital Services consists of leveraging its technologies and government funds to establish a foothold in capturing the growing orbital services market. Lunar Data Services is designed to allow it to provide lunar network services to National Aeronautics and Space Administration and commercial clients. Space Products and Infrastructure includes propulsion systems, navigation systems, engineering services contracts, lunar mobility vehicles, power infrastructure, and human habitation systems.
Graphene has properties almost as if out of a picture book. It is simultaneously strong and flexible, it permeates water but filters gases, it is transparent like glass but at the same time conducts current better than a metal – and on top of it all, it is as thin as can be. Graphene consists just of a single layer of atoms. No surprise that interest in this wonder material (Nobel Prize 2010) explodes. Unfortunately, graphene displays all of these properties only when it is essentially defect free, and this is exactly what limits commercial production today.
Recently, highest-quality graphene films could be produced at copper catalysts. The gist was to melt the copper above 1100° degrees, so that the film growth occurs above the liquid metal surface. Yet, why these extreme conditions favor the formation of defect-free graphene is largely unknown. Not least the high temperatures and gas flows during growth hamper gaining the necessary atomic-level insights. Much hope is therefore placed on methods of AI/machine learning that can accelerate highest-accuracy quantum mechanical predictions to a level that allows so-called molecular dynamics simulations to correctly account for the high mobility of atoms at the liquid copper surface. Up to now, there is very little experience as to the reliability of these AI approaches though.
As an important benchmark, scientists at the Fritz Haber Institute (FHI) predicted the height of a growing graphene layer above the liquid copper catalyst using these novel approaches, and compared it against measurements from high-level synchrotron experiments achieved within a European consortium. The result is striking. The machines were right down to 10-11 m, or within less than one millionth the width of a hair. This high precision, called “sub-Ångstrøm“ in the jargon of the field, demonstrates the impressive capabilities of the novel AI approaches to acquire detailed insights into the microscopic growth process. “Our results nicely show the unprecedented possibilities”, summarizes the leader of the FHI-team Dr. Heenen proudly. “Intriguingly, the data on adsorption height suggest an essentially identical chemical interaction of graphene with solid and liquid copper. This renders the superior synthesis above the liquid metal surface even more mysterious.”
Google has introduced a family of multimodal models, called Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that the most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. Thus, Google believes that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. In this article the authors discuss their approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI.
Google has been rigorously testing all their Gemini models and evaluating their performance on a wide variety of tasks. From natural image, audio and video understanding to mathematical reasoning, Gemini Ultra’s performance exceeds current state-of-the-art results on 30 of the 32 widely-used academic benchmarks used in large language model (LLM) research and development.
With a score of 90.0%, Gemini Ultra is the first model to outperform human experts on MMLU (massive multitask language understanding), which uses a combination of 57 subjects such as math, physics, history, law, medicine and ethics for testing both world knowledge and problem-solving abilities. This new benchmark approach to MMLU enables Gemini to use its reasoning capabilities to think more carefully before answering difficult questions, leading to significant improvements over just using its first impression.
In summary, Gemini surpasses state-of-the-art performance on a range of benchmarks including text and coding. Gemini Ultra also achieves a state-of-the-art score of 59.4% on the new MMMU benchmark, which consists of multimodal tasks spanning different domains requiring deliberate reasoning. With the image benchmarks we tested, Gemini Ultra outperformed previous state-of-the-art models, without assistance from optical character recognition (OCR) systems that extract text from images for further processing. These benchmarks highlight Gemini’s native multimodality and indicate early signs of Gemini's more complex reasoning abilities.
The AGI race has begun. We are building machines that can think and reason. By 2025/26, these machines will outpace many college graduates. By the end of the decade, they will be smarter than any human being; we will have superintelligence, in the true sense of the word. Along the way, national security forces not seen in half a century will be unleashed, and before long, The Project will be on. If we’re lucky, we’ll be in an all-out race with the CCP; if we’re unlucky, an all-out war.
Everyone is now talking about AI, but few have the faintest glimmer of what is about to hit them. Nvidia stock analysts still think 2024 might be close to the peak. Mainstream pundits are stuck on the willful blindness of “it’s just predicting the next word”. They see only hype and business-as-usual; at most they entertain another internet-scale technological change.
Before long, the world will wake up. But right now, there are perhaps a few hundred people, most of them in San Francisco and the AI labs, that have situational awareness. Through whatever peculiar forces of fate, I have found myself amongst them. A few years ago, these people were derided as crazy—but they trusted the trendlines, which allowed them to correctly predict the AI advances of the past few years. Whether these people are also right about the next few years remains to be seen. But these are very smart people and they are the ones building this technology. Perhaps they will be an odd footnote in history, or perhaps they will go down in history like Szilard and Oppenheimer and Teller. If they are seeing the future even close to correctly, we are in for a wild ride. [Leopold Aschenbrenner]
A new report reinforces that the virus is not going away. Even as national institutions struggle to coordinate meaningful trials for possible long COVID treatments, researchers continue to tally the damage. New findings suggest that the disease’s reach isn’t merely long—it’s still growing. Three years after their initial bouts with COVID-19, patients who’d once been hospitalized with the virus remained at “significantly elevated” risk of death or worsening health from long COVID complications, according to a paper published May 30, 2024 in Nature Medicine. Even among those whose initial cases didn’t require a hospital stay, the threat of long COVID and several of its associated issues remained real, the researchers found. And cumulatively, at three years, long COVIDresults in 91 disability-adjusted life years (DALY) per 1,000 people—DALYs being a measure of years lost to poor health or premature death. That is a higher incidence than either heart disease or cancer.
“People are developing new-onset disease as the result of an infection that they had three years ago,” says Dr. Ziyad Al-Aly, a clinical epidemiologist at Washington University in St. Louis and lead author of the study. “It challenges the notion that these viruses are sort of self-contained or that after the acute first phase, they become inconsequential.” With more than 130,000 patients, the study is by far the largest so far to track the progress of the virus over a full three-year period. It expands on work by Al-Aly and others at the two-year mark that found patients had elevated risk for long-COVID-related conditions that included diabetes, lung problems, fatigue, blood clots, and gastrointestinal and musculoskeletal disorders. At three years, Al-Aly tells Fortune,the primary complications among those with mild initial COVID cases were found in the neurological, GI, and pulmonary systems. The persistent risk among those who’d been hospitalized, meanwhile, extended to seven organ systems and included severe conditions such as strokes, heart attacks, heart failure, and even Alzheimer’s disease. The study included nationally recognized researchers Al-Aly and coauthor Dr. Eric Topol, executive vice president and professor of molecular medicine at Scripps Research. It drew from patients within the Veterans Affairs St. Louis Health Care system. As such, the scientists note, the demography skews more male, white, and slightly older than other patient studies might. “The data are encouraging in that there were no new-onset adverse health problems found in the third year after infection,”says Akiko Iwasaki, director of the Center for Infection and Immunity at the Yale University School of Medicine.But Iwasaki, who was not involved in the study, cautioned that some post-acute infection illnesses can turn up years later: “We will need to keep this type of long-term follow-up studies for extended periods.”
More health challenges for hospitalized patients
Perhaps unsurprisingly, those whose initial COVID cases required hospitalization faced the greatest challenges over the course of the three-year study,a grim reminder that interventions like vaccinations and antivirals are critical, Al-Aly says. “The story in hospitalized people is more stark,” the researcher says. “They have greater risk and longer risk horizon, with a burden of disease that is astronomically higher than non-infected people and higher than non-hospitalized individuals. Preventing hospitalization is very important.” The risk of new long COVID complications declined over time for both hospitalized and non-hospitalized patients, the study found. “That’s the good news story,” says Al-Aly.The risk of death, meanwhile, became “insignificant” after the first year among those who didn’t have to go to a hospital—that is, most of us who’ve ever been infected by COVID. For those whose cases required hospitalization, though, the threat of death “remained persistently elevated even in the third year,” the researchers said. That group also faced far greater burden of health—about 90 DALYs per 1,000 people, compared to about 10 DALYs per 1,000 for the non-hospitalized group. (For context, both heart disease and cancer cause about 50 DALYs per 1,000 people.) “The difference in DALYs between the two groups should not be interpreted to mean that people with long COVID from less severe acute disease are not suffering greatly as a result of their long COVID symptoms,” says Dr. David Putrino, director of the Cohen Center for Recovery from Complex Chronic Illness at Mt. Sinai School of Medicine. “It only means that at three years out, they’re experiencing less overtly life-threatening sequelae” than those who initially required hospitalization. The stubborn presence of long COVID reinforces the notion that this is no ordinary virus. Viral persistence, chronic inflammation, and immune system dysfunction are all thought by scientists to come into play, though more study is needed.
What is the outlook for tackling long COVID?
As for the question of what the term long COVID actually means—that is, how long it lasts—the science is still developing, Al-Aly says. He describes long COVID as “this sort of complex web of 80 or more different health problems,” some of which, like stroke or heart disease, could negatively affect patients for the rest of their lives. “We don’t know what we don’t know,” says Al-Aly. “This is only at three years… We don’t know what’s going to happen at 10 years.” The research group is hoping to persuade its funders to continue following the patient cohort for at least that long, he said. For these and other reasons, researchers have led the charge to accelerate the pace of trials for long COVID treatments, the early efforts at which have been criticized for wasting money and critical time. Appearing before a U.S. Senate committee in January, Al-Aly told committee members that at least 20 million people in the country have been hit with long COVID.Globally, that number is estimated to be at least 65 million. That meeting was noteworthy in part because it represented the first Congressional hearing on long COVID since the start of the pandemic. Researchers hope that this latest study will again focus attention on a disease with a long tail and an unknowable future, and perhaps prod the National Institutes of Health to take a bigger swing at the issue. “We need to be much bolder and much more ambitious with our trials,” Al-Aly says. “At the glacial pace that they’re going, we’re unlikely to get any definitive answers for decades to come.”
To protect and rear their young, some insects have transformed wild viruses into tiny biological weapons. If you puncture the ovary of a wasp called Microplitis demolitor, viruses squirt out in vast quantities, shimmering like iridescent blue toothpaste. “It’s very beautiful, and just amazing that there’s so much virus made in there,” says Gaelen Burke, an entomologist at the University of Georgia. M. demolitor is a parasite that lays its eggs in caterpillars, and the particles in its ovaries are “domesticated” viruses that have been tuned to persist harmlessly in wasps and serve their purposes. The virus particles are injected into the caterpillar through the wasp’s stinger, along with the wasp’s own eggs. The viruses then dump their contents into the caterpillar’s cells, delivering genes that are unlike those in a normal virus. Those genes suppress the caterpillar’s immune system and control its development, turning it into a harmless nursery for the wasp’s young.
The insect world is full of species of parasitic wasps that spend their infancy eating other insects alive. And for reasons that scientists don’t fully understand, they have repeatedly adopted and tamed wild, disease-causing viruses and turned them into biological weapons. Half a dozen examples already are described, and new research hints at many more. By studying viruses at different stages of domestication, researchers today are untangling how the process unfolds.
Scientists used a neural network, a type of brain-inspired machine learning algorithm, to sift through large volumes of particle collision data.
For over two decades, the ATLAS particle detector has recorded the highest energy particle collisions in the world within the Large Hadron Collider (LHC) located at CERN, the European Organization for Nuclear Research. Beams of protons are accelerated around the LHC at close to the speed of light, and upon their collision at ATLAS, they produce a cascade of new particles, resulting in over a billion particle interactions per second.
Particle physicists are tasked with mining this massive and growing store of collision data for evidence of undiscovered particles. In particular, they’re searching for particles not included in the Standard Model of particle physics, our current understanding of the universe’s makeup that scientists suspect is incomplete.
As part of the ATLAS collaboration, scientists at the U.S. Department of Energy’s (DOE) Argonne National Laboratory and their colleagues recently used a machine learning approach called anomaly detection to analyze large volumes of ATLAS data. The method has never before been applied to data from a collider experiment. It has the potential to improve the efficiency of the collaboration’s search for something new. The collaboration involves scientists from 172 research organizations.
The team leveraged a brain-inspired type of machine learning algorithm called a neural network to search the data for abnormal features, or anomalies. The technique breaks from more traditional methods of searching for new physics. It is independent of — and therefore unconstrained by — the preconceptions of scientists.
Proving geometric theorems constitutes a hallmark of visual reasoning combining both intuitive and logical skills. Therefore, automated theorem proving of Olympiad-level geometry problems is considered a notable milestone in human-level automated reasoning. The introduction of AlphaGeometry, a neuro-symbolic model trained with 100 million synthetic samples, marked a major breakthrough. It solved 25 of 30 International Mathematical Olympiad (IMO) problems whereas the reported baseline based on Wu’s method solved only ten.
In this paper, the IMO-AG-30 Challenge introduced with AlphaGeometry was revisited, and the researchers found that Wu’s method is surprisingly strong. Wu’s method alone can solve 15 problems, and some of them are not solved by any of the other methods. This leads to two key findings: (i) Combining Wu’s method with the classic synthetic methods of deductive databases and angle, ratio, and distance chasing solves 21 out of 30 methods by just using a CPU-only laptop with a time limit of 5 minutes per problem. Essentially, this classic method solves just 4 problems less than AlphaGeometry and establishes the first fully symbolic baseline, strong enough to rival the performance of an IMO silver medalist. (ii) Wu’s method even solves 2 of the 5 problems that AlphaGeometry failed to solve. Thus, by combining AlphaGeometry with Wu’s method a new state-of-the-art for automated theorem proving on IMO-AG-30, solving 27 out of 30 problems, the first AI method which outperforms an IMO gold medalist is finally achieved.
The COVID-19 pandemic was initiated by the rapid spread of a SARS-CoV-2 strain. Though mainly classified as a respiratory disease, SARS-CoV-2 infects multiple tissues throughout the human body, leading to a wide range of symptoms in patients. To better understand how SARS-CoV-2 affects the proteome from cells with different ontologies, this work generated an infectome atlas of 9 cell models, including cells from brain, blood, digestive system, and adipocyte tissue.
Our data shows that SARS-CoV-2 infection mainly trigger dysregulations on proteins related to cellular structure and energy metabolism. Despite these pivotal processes, heterogeneity of infection was also observed, highlighting many proteins and pathways uniquely dysregulated in one cell type or ontological group. These data have been made searchable online via a tool that will permit future submissions of proteomic data (https://reisdeoliveira.shinyapps.io/Infectome_App/) to enrich and expand this knowledgebase.
NVIDIA recently announced GeForce RTX™ SUPER desktop GPUs for supercharged generative AI performance, new AI laptops from every top manufacturer, and new NVIDIA RTX™-accelerated AI software and tools for both developers and consumers.
Building on decades of PC leadership, with over 100 million of its RTX GPUs driving the AI PC era, NVIDIA is now offering these tools to enhance PC experiences with generative AI: NVIDIA TensorRT™ acceleration of the popular Stable Diffusion XL model for text-to-image workflows, NVIDIA RTX Remix with generative AI texture tools, NVIDIA ACE microservices and more games that use DLSS 3 technology with Frame Generation.
AI Workbench, a unified, easy-to-use toolkit for AI developers, will be available in beta later this month. In addition, NVIDIA TensorRT-LLM (TRT-LLM), an open-source library that accelerates and optimizes inference performance of the latest large language models (LLMs), now supports more pre-optimized models for PCs. Accelerated by TRT-LLM, Chat with RTX, an NVIDIA tech demo also releasing this month, allows AI enthusiasts to interact with their notes, documents and other content.
“Generative AI is the single most significant platform transition in computing history and will transform every industry, including gaming,” said Jensen Huang, founder and CEO of NVIDIA. “With over 100 million RTX AI PCs and workstations, NVIDIA is a massive installed base for developers and gamers to enjoy the magic of generative AI.”
Running generative AI locally on a PC is critical for privacy, latency and cost-sensitive applications. It requires a large installed base of AI-ready systems, as well as the right developer tools to tune and optimize AI models for the PC platform. To meet these needs, NVIDIA is delivering innovations across its full technology stack, driving new experiences and building on the 500+ AI-enabled PC applications and games already accelerated by NVIDIA RTX technology.
This is NVIDIA's first and very important step towards the vision of "LLM as Operating System" - a locally running, heavily optimized AI assistant that can deeply integrate with all your local files, but at the same time preserving privacy. NVIDIA is going local even before OpenAI!
Retroviruses are viruses that multiply by incorporating their genes into the genome of a host cell. If the infected cell is a germ cell, the retrovirus can then be passed on to the next generation as an “endogenous” retrovirus (ERV) and spread as part of the host genome in that host species. In vertebrates, ERVs are ubiquitous and sometimes make up 10 per cent of the host genome. However, most retrovirus integrations are very old, already degraded and therefore inactive – their initial impact on host health has been minimized by millions of years of evolution. A research team led by the Leibniz Institute for Zoo and Wildlife Research (Leibniz-IZW) has now discovered a recent case of retrovirus colonization in a rodent from New Guinea, the white-bellied mosaic-tailed rat. In a PNAS paper they describe this new model of virus integration. The observations on this process will help to improve our understanding how retroviruses rewrite host genomes.
Retroviruses, such as the pathogen responsible for AIDS (HIV-1), integrate into the genome of the host cells they infect during their life cycle. When this happens in the germline (egg cells or cells that produce sperm) of the host, the retrovirus can actually become a gene of the host itself. This process is apparently common, as up to 10 percent of the genomes of most vertebrates consist of the remnants of such ancient infections. One of the best studied models of this process is the koala retrovirus (KoRV), which is currently colonizing the koala genome. „What happens to the virus and the host during this process of genome colonization we do not know, as most such events occurred millions of years ago and we only see the leftover ‘fossils’ of the retrovirus”, says Prof Alex Greenwood, head of the Leibniz-IZW Department of Wildlife Diseases. “Nor do we know what the host suffered health-wise during the infection process. The koala retrovirus (KoRV) is one of the few models of this process that occurs in real time and where we can observe the effects of genome colonization on the host animal.”
There is now some evidence that KoRV-related viruses are circulating in rodents and bats in Papua New Guinea and Indonesia. A group led by Greenwood and Dr Saba Mottaghinia, former PhD student in Greenwood’s department, analyzed 278 samples from seven bat and one rodent family endemic to Australia and New Guinea. They discovered a retrovirus that is currently colonizing the genome of an endemic rodent from New Guinea, the white-bellied mosaic-tailed rat (Melomys leucogaster). This is only the second example from the Australo-Papuan region, after KoRV, of a retrovirus that has colonized a genome while retaining a functional viral life cycle. The gibbon ape leukaemia viruses (GALV), a group of viruses discovered in gibbons and woolly monkeys at a research facility in Thailand in the 1960s, are very closely related to KoRV. This is a curious and surprising relationship, as there is a geographical barrier, known as the Wallace lineage, which separates the fauna of Southeast Asia from Indonesia, Papua New Guinea and Australia's fauna. However, there is evidence that the gibbons and woolly monkeys at the research facility in Thailand have been infected with viruses from Papua New Guinea. „The discovery of GALV-like viruses in rodents and bats in Indonesian and Australian rodents and bats from New Guinea suggests that these viruses, and possibly also KoRV, originated in New Guinea”, says Greenwood, who initiated the research project funded by the German Research Foundation.
The Leibniz-IZW team, together with scientists from the Charité, the Robert Koch Institute, the Max Delbrück Center, the University of Nicosia, California State University Fullerton, the South Australian Museum and Museum Victoria, examined 278 bat and rodent samples from Australia and New Guinea for KoRV and GALV-like viruses. They detected a GALV, the Woolly Monkey Virus (WMV) in a population of the white-bellied mosaic-tailed rat, an endemic rodent from New Guinea. In five of the rats from two New Guinea collection sites, WMV was integrated into the genome at the same location, indicating that it has spread as a gene and not by infection, i.e. it has become part of the genome. However, in other white-bellied mosaic-tailed rat populations the virus was absent – similar to KoRV in koalas, where all koalas in northern Australia have KoRV in their genome, whereas there are koalas in southern Australia that do not have intact KoRV. The virus, now called the “complete Melomys woolly monkey virus” (cMWMV), was able to infect cell lines, produce new viral progeny, was visible by electron microscopy as viral particles that detached from the cell membrane, and was even sensitive to the antiretroviral drug AZT.
„The virus has all the characteristics of an exogenous infectious retrovirus, but is endogenous. It is probably a very recent colonization event, much younger than KoRV", says Dr Saba Mottaghinia, the lead author of the paper. The results suggest that cMWMV is a new model for retroviral colonization of the host genome that occurs in real time, as in KoRV, and they also suggest that GALVs like WMV originated in the diverse fauna of New Guinea. The discoveries in New Guinea have certainly not been exhausted. “There are hundreds of mammalian species from this region that have not yet been studied, suggesting that many more viruses and models exist in this region”, says Greenwood.
The first surgery demonstration in space performed by a robot — controlled by surgeons on Earth — has marked a technological milestone for long-term spaceflight.
The robot, known as spaceMIRA — which stands for Miniaturized In Vivo Robotic Assistant — performed several operations on simulated tissue at the orbiting laboratory while remotely operated by surgeons from approximately 250 miles (400 kilometers) below in Lincoln, Nebraska.
The milestone is a step forward in developing technology that could have implications not just for successful long-term human space travel, where surgical emergencies could happen, but also for establishing access to medical care in remote areas on Earth.
The interior of black holes remains a conundrum for science. In 1916, German physicist Karl Schwarzschild outlined a solution to Albert Einstein's equations of general relativity, in which the center of a black hole consists of a so-called singularity, a point at which space and time no longer exist. Here, the theory goes, all physical laws, including Einstein's general theory of relativity, no longer apply; the principle of causality is suspended.
This constitutes a great nuisance for science—after all, it means that no information can escape from a black hole beyond the so-called event horizon. This could be a reason why Schwarzschild's solution did not attract much attention outside the theoretical realm—that is, until the first candidate for a black hole was discovered in 1971, followed by the discovery of the black hole in the center of our Milky Way in the 2000s, and finally the first image of a black hole, captured by the Event Horizon Telescope Collaboration in 2019.
In 2001, Pawel Mazur and Emil Mottola proposed a different solution to Einstein's field equations that led to objects that they called gravitational condensate stars, or gravastars. Contrary to black holes, gravastars have several advantages from a theoretical astrophysics perspective.
On the one hand, they are almost as compact as black holes and also exhibit a gravity at their surface that is essentially as strong as that of a black hole, hence resembling a black hole for all practical purposes. On the other hand, gravastars do not have an event horizon, that is, a boundary from within which no information can be sent out, and their core does not contain a singularity.
Instead, the center of a gravastar is made up of an exotic (dark) energy that exerts a negative pressure to the enormous gravitational force compressing the star. The surface of a gravastar is represented by a wafer-thin skin of ordinary matter, the thickness of which approaches zero.
Theoretical physicists Daniel Jampolski and Prof. Luciano Rezzolla of Goethe University Frankfurt have now presented a solution to the field equations of general relativity that describes the existence of a gravastar inside another gravastar. They have given this hypothetical celestial object the name "nestar" (from the English "nested"). The study is published in Classical and Quantum Gravity.
As OpenAI and other model developers build more capable AI systems, the potential for both beneficial and harmful uses of AI will grow. One potentially harmful use, highlighted by researchers and policymakers, is the ability for AI systems to assist malicious actors in creating biological threats (e.g., see White House 2023, Lovelace 2022, Sandbrink 2023). In one discussed hypothetical example, a malicious actor might use a highly-capable model to develop a step-by-step protocol, troubleshoot wet-lab procedures, or even autonomously execute steps of the biothreat creation process when given access to tools like cloud labs (see Carter et al., 2023). However, assessing the viability of such hypothetical examples was limited by insufficient evaluations and data.
Following OpenAI's recently shared Preparedness Framework, they are developing methodologies to empirically evaluate these types of risks, in order to understand both where AI models are today and where they might be in the future. Now, OpenAI details a new evaluation which could help serve as one potential “tripwire” signaling the need for caution and further testing of biological misuse potential. This evaluation aims to measure whether models could meaningfully increase malicious actors’ access to dangerous information about biological threat creation, compared to the baseline of existing resources (i.e., the internet).
To evaluate this, OpenAI conducted a study with 100 human participants, comprising (a) 50 biology experts with PhDs and professional wet lab experience and (b) 50 student-level participants, with at least one university-level course in biology. Each group of participants was randomly assigned to either a control group, which only had access to the internet, or a treatment group, which had access to GPT-4 in addition to the internet. Each participant was then asked to complete a set of tasks covering aspects of the end-to-end process for biological threat creation.A[A]
Findings:
This study assessed uplifts in performance for participants with access to GPT-4 across five metrics (accuracy, completeness, innovation, time taken, and self-rated difficulty) and five stages in the biological threat creation process (ideation, acquisition, magnification, formulation, and release). They found mild uplifts in accuracy and completeness for those with access to the language model. Specifically, on a 10-point scale measuring accuracy of responses, they observed a mean score increase of 0.88 for experts and 0.25 for students compared to the internet-only baseline, and similar uplifts for completeness (0.82 for experts and 0.41 for students). However, the obtained effect sizes were not large enough to be statistically significant, and this study highlighted the need for more research around what performance thresholds indicate a meaningful increase in risk. Moreover, OpenAI notes that information access alone is insufficient to create a biological threat, and that this evaluation does not test for success in the physical construction of the threats.
An experimental gene therapy has restored the hearing of a child who was born deaf, the pharmaceutical giant Eli Lilly said in a statement. The child, who was 11 years old at the time the therapy was administered, experienced restored hearing within 30 days of treatment, Eli Lilly said in a release. The child participating in the clinical study was the first individual in the U.S. to receive the therapy for a genetic form of hearing loss, the company said.
The gene therapy, AK-OTOF, is being developed for the treatment of hearing loss due to mutations in the otoferlin gene. added AK-OTOF to its portfolio through its , a genetic-medicine company focused on inner-ear conditions.
Doctors and scientists have been pursuing gene therapy for hearing loss for more than 20 years, and “these initial results show that it may restore hearing better than many thought possible,” Dr. John Germiller, director of clinical research in the otolaryngology department at the Children’s Hospital of Philadelphia and principal investigator in the clinical trial, said in a statement.
The child’s hearing was restored across all tested frequencies and was within a normal range at some frequencies at 30 days after the treatment, Eli Lilly said.
Children with hearing problems due to otoferlin gene mutations are often born with profound hearing loss, but most of them have not had the genetic testing needed for a definitive diagnosis, Dr. Oliver Haag, head of otolaryngology at Sant Joan de Deu Hospital in Barcelona and an Akouos study investigator, said in a statement.
To get content containing either thought or leadership enter:
To get content containing both thought and leadership enter:
To get content containing the expression thought leadership enter:
You can enter several keywords and you can refine them whenever you want. Our suggestion engine uses more signals but entering a few keywords here will rapidly give you great content to curate.
Your new post is loading...
