Amazing Science
1.0M views | +30 today
Scooped by Dr. Stefan Gruenwald
onto Amazing Science!

Retina cells 3D printed for the first time

Retina cells 3D printed for the first time | Amazing Science |

The ability to arrange cells into highly defined patterns and structures has recently elevated the use of 3D printing in the biomedical sciences to create cell-based structures for use in regenerative medicine.

A group of researchers from the UK have used 3-D biomedical printing to successfully print new eye cells, making it the first time the technology has been used successfully to print mature central nervous system cells. The breakthrough could lead to the production of artificial tissue grafts made from the variety of cells found in the human retina and may aid in the search to cure blindness.

Experts at the University of Cambridge printed two types of cells - ganglion cells and glial cells - derived from adult rat retinas. Ganglion cells transmit information from the eye to parts of the brain, while glial cells provide support and protection for neurons.

Co-authors of the study Professor Keith Martin and Dr Barbara Lorber, from the John van Geest Centre for Brain Repair, University of Cambridge, said: "The loss of nerve cells in the retina is a feature of many blinding eye diseases. The retina is an exquisitely organised structure where the precise arrangement of cells in relation to one another is critical for effective visual function".

In their study, the researchers used a single nozzle piezoelectric inkjet printer that ejected the cells through a sub-millimetre diameter nozzle when a specific electrical pulse was applied. The driving waveform was defined by a PC-driven generator. "We plan to extend this study to print other cells of the retina and to investigate if light-sensitive photoreceptors can be successfully printed using inkjet technology. In addition, we would like to further develop our printing process to be suitable for commercial, multi-nozzle print heads," Professor Martin concluded. His goal is to make living tissues using multiple nozzles so that different types of cells could be printed from different nozzles at the same time.

The study has been detailed in a paper published in Biofabrication.

No comment yet.
Amazing Science
Amazing science facts - 3D_printing • aging • AI • anthropology • art • astronomy • bigdata • bioinformatics • biology • biotech • chemistry • computers • cosmology • education • environment • evolution • future • genetics • genomics • geosciences • green_energy • language • map • material_science • math • med • medicine • microscopy • nanotech • neuroscience • paleontology • photography • photonics • physics • postings • robotics • science • technology • video
Your new post is loading...
Scooped by Dr. Stefan Gruenwald!

Unlimited, at-home coronavirus testing for your organization

Unlimited, at-home coronavirus testing for your organization | Amazing Science |



Toll Free:1-800-605-8422  FREE
Regular Line:1-858-345-4817


This newsletter is aggregated from over 1450 news sources:


All my Tweets and Scoop.It! posts sorted and searchable:

archived twitter feed


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.


You can also type your own query: e.g., you are looking for articles involving "DNA" as a keyword


Or CLICK on the little FUNNEL symbol at the





• 3D-printing • aging • AI • anthropology • art • astronomy • bigdata • bioinformatics • biology • biotech • chemistry • computers • cosmology • education • environment • evolution • future • genetics • genomics • geosciencesgreen-energy • history • language • mapmaterial-science • math • med • medicine • microscopymost-reads • nanotech • neuroscience • paleontology • photography • photonics • physics • postings • robotics • science • technology • video 



topthcvapestore's comment, July 1, 2021 2:05 AM
Traluma's comment, July 10, 2021 8:07 AM
modricscouk's comment, March 24, 1:37 AM
Scooped by Dr. Stefan Gruenwald!

The end of art or a new beginning? DALL·E 2 Will Disrupt Art Deeper Than Photography Did

The end of art or a new beginning? DALL·E 2 Will Disrupt Art Deeper Than Photography Did | Amazing Science |

Are we going to soon witness the creative genius of a virtual Picasso or Leonardo da Vinci? We could be on the way with the latest trend in the AI field: visual generative models. While systems like GPT-3 create text from text, others like DALL·E 2 — a wordplay between Spanish painter Salvador Dalí and Pixar’s cute robot WALL·E — can create visual art from just words and sentences. As you can see from the cover image of this article, this tech gives a new meaning to the idiom “an image is worth more than a thousand words.”


OpenAI, the company behind DALL·E 2, announced the model in April. In just a couple of months, it has sounded all the alarms that AI is getting too close to human-level visual creativity. It can create anything you can imagine — and anything you can’t — from a single sentence. 


And OpenAI isn’t alone. Just a month later, Google Brain announced Imagen, a similar — and even slightly better — model. And now, another team from Google Brain released yet another model called Parti. And that’s without mentioning all the projects coming from low-budget research groups that don’t count on the resources of the Googles and OpenAIs of the world. Some examples are brand new research labs like Midjourney, free apps like Dream, open-source alternatives like Hugging Face’s DALL·E mini (now called Craiyon), or the collab notebooks (e.g. disco diffusion) that originate from the work of a few independent researchers like Katherine Crowson and Ryan Murdock, who started the movement among digital artists.


Now that the tech is proven and the uncertainty removed, the whole AI community is in a race to build the next best text-to-image model. And they aren’t stopping anytime soon. This will have important consequences. Today, I’ll focus on how this technology will disrupt the visual arts. DALL·E 2, the main star of today’s issue, will accompany us on this journey.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

AI and cryo-electron tomography: Towards Visual Proteomics at High Resolution

AI and cryo-electron tomography: Towards Visual Proteomics at High Resolution | Amazing Science |

Traditionally, structural biologists approach the complexity of cellular proteomes in a reductionist manner. Proteomes are fractionated, their molecular components purified and studied one-by-one using the experimental methods for structure determination at their disposal. Visual proteomics aims at obtaining a holistic picture of cellular proteomes by studying them in situ, ideally in unperturbed cellular environments. The method that enables doing this at highest resolution is cryo-electron tomography. It allows to visualize cellular landscapes with molecular resolution generating maps or atlases revealing the interaction networks which underlie cellular functions in health and in disease states. Current implementations of cryo-ET do not yet realize the full potential of the method in terms of resolution and interpretability. To this end, further improvements in technology and methodology are needed. This review describes the state of the art as well as measures which we expect will help overcoming current limitations.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

NASA Reconnaissance Orbiter Spots Rocket Impact Site of Unknown Origin on Moon

NASA Reconnaissance Orbiter Spots Rocket Impact Site of Unknown Origin on Moon | Amazing Science |

Astronomers discovered a rocket body heading toward a lunar collision late last year. Impact occurred March 4, with NASA's Lunar Reconnaissance Orbiter later spotting the resulting crater. Surprisingly the crater is actually two craters, an eastern crater (18-meter diameter, about 19.5 yards) superimposed on a western crater (16-meter diameter, about 17.5 yards). A rocket body impacted the Moon on March 4, 2022, near Hertzsprung crater, creating a double crater roughly 28 meters wide in the longest dimension.


The double crater was completely unexpected and may indicate that the rocket body had large masses at each end. Typically a spent rocket has mass concentrated at the motor end; the rest of the rocket stage mainly consists of an empty fuel tank. Since the origin of the rocket body remains uncertain, the double nature of the crater may indicate its identity. The crater formed (5.226 degrees north, 234.486 degrees east, 1,863 meters elevation) in a complex area where the impact of ejecta from the Orientale basin event overlies the degraded northeast rim of Hertzsprung basin (536 kilometers diameter). The new crater is not visible in this view, but its location is indicated by the white arrow. LROC WAC mosaic, 110 kilometers width.


No other rocket body so far that crashed on the Moon has created double craters. The four Apollo SIV-B craters were somewhat irregular in outline (Apollos 13, 14, 15, 17) and were substantially larger (greater than 35 meters, about 38 yards) than each of the double craters. The maximum width (29 meters, about 31.7 yards) of the double crater of the mystery rocket body was near that of the S-IVBs.

LRO is managed by NASA's Goddard Space Flight Center in Greenbelt, Maryland, for the Science Mission Directorate at NASA Headquarters in Washington. Launched on June 18, 2009, LRO has collected a treasure trove of data with its seven powerful instruments, making an invaluable contribution to our knowledge about the Moon. NASA is returning to the Moon with commercial and international partners to expand human presence in space and bring back new knowledge and opportunities.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Endothelial cell heterogeneity based on pig cell landscape at single-cell level

Endothelial cell heterogeneity based on pig cell landscape at single-cell level | Amazing Science |

Pigs are valuable large animal models for biomedical and genetic research, but insights into the tissue- and cell-type-specific transcriptome and heterogeneity remain limited. By leveraging single-cell RNA sequencing, scientists now generated a multiple-organ single-cell transcriptomic map containing over 200,000 pig cells from 20 tissues/organs. They were able to comprehensively characterize the heterogeneity of cells in tissues and to identify 234 cell clusters, representing 58 major cell types. In-depth integrative analysis of endothelial cells reveals a high degree of heterogeneity. They also identified several functionally distinct endothelial cell phenotypes, including an endothelial to mesenchymal transition subtype in adipose tissues. Intercellular communication analysis predicts tissue- and cell type-specific crosstalk between endothelial cells and other cell types through the VEGF, PDGF, TGF-β, and BMP pathways. Regulon analysis of single-cell transcriptome of microglia in pig and 12 other species further identifies MEF2C as an evolutionally conserved regulon in the microglia.


This important work describes the landscape of single-cell transcriptomes within diverse pig organs and identifies the heterogeneity of endothelial cells and evolutionally conserved regulon in microglia.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Intelligent Toilet Offers Remote Patient Monitoring for Gastrointestinal Health

Researchers at Duke University are developing an artificial intelligence tool for toilets that would help providers improve care management for patients with gastrointestinal issues through remote patient monitoring.  The tool, which can be installed in the pipes of a toilet and analyzes stool samples, has the potential to improve treatment of chronic gastrointestinal issues like inflammatory bowel disease or irritable bowel syndrome, according to a press release. When a patient flushes the toilet, the mHealth platform photographs the stool as it moves through the pipes. That data is sent to a gastroenterologist, who can analyze the data for evidence of chronic issues.


A study conducted by Duke University researchers found that the platform had an 85.1 percent accuracy rate on stool form classification and a 76.3 percent accuracy rate on detection of gross blood. “Typically, gastroenterologists have to rely on patient self-reported information about their stool to help determine the cause of their gastrointestinal health issues, which can be very unreliable,” Deborah Fisher, MD, an associate professor of medicine at Duke and one of the study’s lead authors, said in the press release.


“Patients often can’t remember what their stool looks like or how often they have a bowel movement, which is part of the standard monitoring process,” she said. “The smart toilet technology will allow us to gather the long-term information needed to make a more accurate and timely diagnosis of chronic gastrointestinal problems.”


This remote patient monitoring platform has the potential to benefit both patients and healthcare providers. Patients won’t need to do anything out of the ordinary, and gastroenterologists won’t have to try and diagnose based on a patient’s description or recollection. “An IBD flare-up could be diagnosed using the smart toilet and the patient’s response to treatment could be monitored with the technology,” Sonia Grego, PhD, a lead researcher on the study and founding director of the Duke Smart Toilet Lab, said in the release. “This could be especially useful for patients who live in long-term care facilities who may not be able to report their conditions and could help improve initial diagnosis of acute conditions,”  The program makes use of something that nearly every American has at home. 


It’s not the first study to target the toilet. Back in 2019, researchers at the Rochester Institute of Technology developed a  sensor-embedded toilet seat. The toilet seat was designed to allow care providers to remotely monitor a patient’s weight, heart rate, blood pressure, blood oxygenation levels, and stroke volume.  Similar to the AI tool for toilet pipes, RIT’s smart toilet seat presents the ability to catch patients’ symptoms early enough before they could lead to something more severe.


With gastroenterology appearing on Doximity’s list of specialties that are least engaged in telehealth, this artificial intelligence software, once available to the public, could change minds. Gastroenterology usually requires in-person and hands-on interaction but using remote monitoring in this way might prove to be even more beneficial.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Giant Filamentous Bacteria of 1 Millimeter Length Found in Mangrove Swamps Challenging Traditional Concepts of Biology

Giant Filamentous Bacteria of 1 Millimeter Length Found in Mangrove Swamps Challenging Traditional Concepts of Biology | Amazing Science |
Researchers describe a “’macro’ microbe” – a giant filamentous bacterium composed of a single cell discovered in the mangroves of Guadeloupe.


At first glance, the slightly murky waters in the tube look like a scoop of stormwater, complete with leaves, debris, and even lighter threads in the mix. But in the Petri dish, the thin vermicelli-like threads floating delicately above the leaf debris are revealed to be single bacterial cells, visible to the naked eye. The unusual size is notable because bacteria aren't usually visible without the assistance of microscope. "It's 5,000 times bigger than most bacteria. To put it into context, it would be like a human encountering another human as tall as Mount Everest," said Jean-Marie Volland, a scientist with joint appointments at the U.S. Department of Energy (DOE) Joint Genome Institute (JGI), a DOE Office of Science User Facility located at Lawrence Berkeley National Laboratory (Berkeley Lab) and the Laboratory for Research in Complex Systems (LRC) in Menlo Park, Calif. In the June 24, 2022, issue of the journal Science, Volland and colleagues, including researchers at the JGI and Berkeley Lab, LRC, and at the Université des Antilles in Guadeloupe, described the morphological and genomic features of this giant filamentous bacterium, along with its life cycle.


For most bacteria, their DNA floats freely within the cytoplasm of their cells. This newly discovered species of bacteria keeps its DNA more organized. "The big surprise of the project was to realize that these genome copies that are spread throughout the whole cell are actually contained within a structure that has a membrane," Volland said. "And this is very unexpected for a bacterium."


Strange Encounters in the Mangroves

The bacterium itself was discovered by Olivier Gros, a marine biology professor at the Université des Antilles in Guadeloupe, in 2009. Gros' research focuses on marine mangrove systems, and he was looking for sulfur-oxidizing symbionts in sulfur-rich mangrove sediments not far from his lab when he first encountered the bacteria. "When I saw them, I thought, 'Strange,'" he said. "In the beginning I thought it was just something curious, some white filaments that needed to be attached to something in the sediment like a leaf." The lab conducted some microscopy studies over the next couple of years, and realized it was a sulfur-oxidizing prokaryote.


Silvina Gonzalez-Rizzo, an associate professor of molecular biology at the Université des Antilles and a co-first author on the study, performed the 16S rRNA gene sequencing to identify and classify the prokaryote. "I thought they were eukaryotes; I didn't think they were bacteria because they were so big with seemingly a lot of filaments," she recalled of her first impression. "We realized they were unique because it looked like a single cell. The fact that they were a 'macro' microbe was fascinating!"


"She understood that it was a bacterium belonging to the genus Thiomargarita," Gros noted. She named it Ca. Thiomargarita magnifica. "Magnifica because magnus in Latin means big and I think it's gorgeous like the French word magnifique," Gonzalez-Rizzo explained. "This kind of discovery opens new questions about bacterial morphotypes that have never been studied before."


Characterizing the Giant Bacterium

Volland got involved with the giant Thiomargarita bacteria when he returned to the Gros lab as a postdoctoral fellow. When he applied to the discovery-based position at the LRC that would see him working at the JGI, Gros allowed him to continue research on the project.


At the JGI, Volland began studying Ca. T. magnifica in Tanja Woyke's Single Cells Group to better understand what this sulfur-oxidizing, carbon fixing bacterium was doing in the mangroves. "Mangroves and their microbiomes are important ecosystems for carbon cycling. If you look at the space that they occupy on a global scale, it's less than 1% of the coastal area worldwide. But when you then look at carbon storage, you'll find that they contribute 10-15% of the carbon stored in coastal sediments," said Woyke, who also heads the JGI's Microbial Program and is one of the article's senior authors. The team was also compelled to study these large bacteria in light of their potential interactions with other microorganisms. "We started this project under the JGI's strategic thrust of inter-organismal interactions, because large sulfur bacteria have been shown to be hot spots for symbionts," Woyke said. "Yet the project took us into a very different direction," she added.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

The octopus’ brain and the human brain share the same 'jumping genes' - LINE elements

The octopus’ brain and the human brain share the same 'jumping genes' - LINE elements | Amazing Science |

The octopus is an exceptional organism with an extremely complex brain and cognitive abilities that are unique among invertebrates. So much so that in some ways it has more in common with vertebrates than with invertebrates. The neural and cognitive complexity of these animals could originate from a molecular analogy with the human brain, as discovered by a research paper recently published in BMC Biology and coordinated by Remo Sanges from SISSA of Trieste and by Graziano Fiorito from Stazione Zoologica Anton Dohrn of Naples. The research shows that the same 'jumping genes' are active both in the human brain and in the brain of two species, Octopus vulgaris, the common octopus, and Octopus bimaculoides, the Californian octopus. A discovery that could help us understand the secret of the intelligence of these fascinating organisms.


Sequencing the human genome revealed as early as 2001 that over 45% of it is composed by sequences called transposons, so-called 'jumping genes' that, through molecular copy-and-paste or cut-and-paste mechanisms, can 'move' from one point to another of an individual's genome, shuffling or duplicating. In most cases, these mobile elements remain silent: they have no visible effects and have lost their ability to move. Some are inactive because they have, over generations, accumulated mutations; others are intact, but blocked by cellular defense mechanisms. From an evolutionary point of view even these fragments and broken copies of transposons can still be useful, as 'raw matter' that evolution can sculpt.


Among these mobile elements, the most relevant are those belonging to the so-called LINE (Long Interspersed Nuclear Elements) family, found in a hundred copies in the human genome and still potentially active. It has been traditionally though that LINEs' activity was just a vestige of the past, a remnant of the evolutionary processes that involved these mobile elements, but in recent years new evidence emerged showing that their activity is finely regulated in the brain. There are many scientists who believe that LINE transposons are associated with cognitive abilities such as learning and memory: they are particularly active in the hippocampus, the most important structure of our brain for the neural control of learning processes.


The octopus' genome, like ours, is rich in 'jumping genes', most of which are inactive. Focusing on the transposons still capable of copy-and-paste, the researchers identified an element of the LINE family in parts of the brain crucial for the cognitive abilities of these animals. The discovery, the result of the collaboration between Scuola Internazionale Superiore di Studi Avanzati, Stazione Zoologica Anton Dohrn and Istituto Italiano di Tecnologia, was made possible thanks to next generation sequencing techniques, which were used to analyze the molecular composition of the genes active in the nervous system of the octopus.


"The discovery of an element of the LINE family, active in the brain of the two octopuses species, is very significant because it adds support to the idea that these elements have a specific function that goes beyond copy-and-paste," explains Remo Sanges, director of the Computational Genomics laboratory at SISSA, who started working at this project when he was a researcher at Stazione Zoologica Anton Dohrn of Naples. The study, published in BMC Biology, was carried out by an international team with more than twenty researchers from all over the world.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Rapid, scalable assessment of SARS-CoV-2 (CoVid19) cellular immunity by whole-blood PCR

Rapid, scalable assessment of SARS-CoV-2 (CoVid19) cellular immunity by whole-blood PCR | Amazing Science |

Fast, high-throughput methods for measuring the level and duration of protective immune responses to SARS-CoV-2 are needed to anticipate the risk of breakthrough infections. A research team now reports the development of two quantitative PCR assays for SARS-CoV-2-specific T cell activation. The assays are rapid, internally normalized and probe-based: qTACT requires RNA extraction and dqTACT avoids sample preparation steps. Both assays rely on the quantification of CXCL10 messenger RNA, a chemokine whose expression is strongly correlated with activation of antigen-specific T cells. On re-stimulation of whole-blood cells with SARS-CoV-2 viral antigens, viral-specific T cells secrete IFN-γ, which stimulates monocytes to produce CXCL10. CXCL10 mRNA can thus serve as a proxy to quantify cellular immunity. These types of assays may allow large-scale monitoring of the magnitude and duration of functional T cell immunity to SARS-CoV-2, thus helping to prioritize revaccination strategies in vulnerable populations. The T cell response to SARS-CoV-2 is detected using a PCR assay on whole blood.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Ionizing Black Hole “Atoms” Formed by a Cloud of Ultralight Bosons Around a Rotating Black Hole

Ionizing Black Hole “Atoms” Formed by a Cloud of Ultralight Bosons Around a Rotating Black Hole | Amazing Science |
Distinctive features of gravitational-wave signals from black hole mergers could reveal the existence of long-sought ultralight bosons.


For decades, physicists have speculated about the existence of ultralight bosons. These hypothetical particles have the potential to solve myriad puzzles in physics, from the strong charge-parity problem to the dark matter mystery. But most proposals for detecting ultralight bosons assume a large particle abundance or a particular coupling to the standard model. Now, John Stout of Harvard University and his colleagues suggest a new detection strategy, based on gravitational-wave observations, that doesn’t depend on those assumptions [1].


The researchers build on previous work by team member Daniel Baumann of the University of Amsterdam (see Synopsis: Black Holes Could Reveal New Ultralight Particles). He and his colleagues used the fact that a cloud of ultralight bosons around a rotating black hole would occupy specific allowed orbits—like an electron orbiting a hydrogen nucleus. They found that if this black hole “atom” were involved in a merger with another black hole at certain points during the inspiral, the black holes’ orbits would become resonant with an allowed orbital transition of the boson cloud. Like an electron excited to a higher energy state by a photon, the cloud would be boosted to higher allowed orbits by the spiraling black holes.


In their new proposal, Stout and his colleagues focus on a later period of the inspiral, when the orbital speed is large enough to “ionize” the ultralight bosons, ripping them from the black holes’ gravitational grasp. They find that the resulting gravitational-wave signal should be significantly altered by this interaction, not only at discrete resonance points but continuously once the binary inspiral reaches the ionization threshold. The gravitational-wave frequencies at which this ionization occurs should be observable by future gravitational-wave detectors such as LISA.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

» Rethinking the rabies vaccine

» Rethinking the rabies vaccine | Amazing Science |

Rabies virus kills a shocking 59,000 people each year, many of them children. Some victims, especially kids, don’t realize they’ve been exposed until it is too late. For others, the intense rabies treatment regimen is out of the question: treatment is not widely available and the average $3,800 expense poses unthinkable economic burden for most people around the world.


Rabies vaccines, rather than treatments, are much more affordable and easier to administer. But those vaccines also come with a massive downside: “Rabies vaccines don’t provide lifelong protection. You have to get your pets boosted every year to three years,” says LJI Professor Erica Ollmann Saphire, Ph.D.. “Right now, rabies vaccines for humans and domestic animals are made from killed virus. But this inactivation process can cause the molecules to become misshapen—so these vaccines aren’t showing the right form to the immune system. If we made a better shaped, better structured vaccine, would immunity last longer?”


Saphire and her team, in collaboration with a team led by Institut Pasteur Professor Hervé Bourhy, DVM, PhD., may have discovered the path to better vaccine design. In a new study, published in Science Advances, the researchers share one of the first high-resolution looks at the rabies virus glycoprotein in its vulnerable “trimeric” form.


“The rabies glycoprotein is the only protein that rabies expresses on its surface, which means it is going to be the major target of neutralizing antibodies during an infection,” says LJI Postdoctoral Fellow Heather Callaway, Ph.D., who serves as the study’s first author.


“Rabies is the most lethal virus we know. It is so much a part of our history—we’ve lived with its specter for hundreds of years,” adds Saphire, who also serves as LJI’s President and CEO. “Yet scientists have never observed the organization of its surface molecule. It is important to understand that structure to make more effective vaccines and treatments—and to understand how rabies and other viruses like it enter cells.”


“This understanding is crucial for designing better vaccines and for developing treatments to cure rabid patients,” says Bourhy.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

ESA/JAXA spacecraft set for second close encounter with Mercury next week

ESA/JAXA spacecraft set for second close encounter with Mercury next week | Amazing Science |

The joint European-Japanese BepiColombo mission is gearing up for its next close flyby of Mercury. This will be the second of six Mercury flybys overall the spacecraft will be making during its seven-year cruise to the least explored planet of the inner Solar System. The closest approach of about 200 km altitude over the surface of Mercury is at 11:44 CEST on Thursday, June 23, the mission team tweeted on Thursday. Launched on 20 October 2018, BepiColombo is a joint mission between the European Space Agency (ESA) and the Japan Aerospace Exploration Agency (JAXA). It is the second mission ever to orbit Mercury and the most complex one.


BepiColombo's first close flyby of the planet took place on 1-2 October 2021. During the October gravity assist manoeuvre, the spacecraft was at an altitude of 199 km from the planet's surface. The third Mercury flyby will take place on June 20, 2023. Overall, the spacecraft will make use of nine planetary flybys - one at Earth, two at Venus, and six at Mercury, together with the spacecraft's solar electric propulsion system - to help steer on course for Mercury orbit in 2025.


The mission comprises two scientific orbiters: the ESA-led Mercury Planetary Orbiter (MPO) and JAXA's Mercury Magnetospheric Orbiter (MMO) to study all aspects of the planet - from the structure and dynamics of its magnetosphere and how it interacts with the solar wind, to its internal structure with its large iron core, and the origin of the planet's magnetic field.


The European-Japanese BepiColombo mission is expected to begin its science operations in February 2026.

No comment yet.
Rescooped by Dr. Stefan Gruenwald from Virus World!

The Largest Ever Series of Phage Therapy Case Studies Shows a Success Rate of Over 50%

The Largest Ever Series of Phage Therapy Case Studies Shows a Success Rate of Over 50% | Amazing Science |

The number of reported cases using viruses to treat deadly Mycobacterium infections just went up by a factor of five. In a new study, a team led by researchers from the University of Pittsburgh and the University of California San Diego report 20 new case studies on the use of the experimental treatment, showing the therapy’s success in more than half of the patients. The number of reported cases using viruses to treat deadly Mycobacterium infections just went up by a factor of five. In a new study, a team led by researchers from the University of Pittsburgh and the University of California San Diego report 20 new case studies on the use of the experimental treatment, showing the therapy’s success in more than half of the patients.  It’s the largest ever set of published case studies for therapy using bacteria-killing viruses known as bacteriophages, providing unprecedented detail on their use to treat dire infections while laying the groundwork for a future clinical trial. “Some of those are spectacular outcomes, and others are complicated,” said Graham Hatfull, the Eberly Family Professor of Biotechnology in the Kenneth P. Dietrich School of Arts and Sciences at the University of Pittsburgh. “But when we do 20 cases, it becomes much more compelling that the phages are contributing to favorable outcomes — and in patients who have no other alternatives.” Each patient treated in the study was infected with one or more strains of Mycobacterium, a group of bacteria that can cause deadly, treatment-resistant infections in those with compromised immune systems or with the lung disorder cystic fibrosis.


In 2019, Hatfull led a team showing the first successful use of phages to treat one of these infections. “For clinicians, these are really a nightmare: They’re not as common as some other types of infections, but they’re amongst some of the most difficult to treat with antibiotics,” said Hatfull. “And especially when you take these antibiotics over extended periods of time, they’re toxic or not very well-tolerated.”  Last month, the University of Pittsburgh researchers were involved in two new case studies successfully treating patients in collaboration with colleagues at National Jewish Health and Harvard University. But those reports represent only a fraction of the cases the team has been involved in behind the scenes. Since 2019, Hatfull and his lab have fielded requests from more than 200 clinicians looking for treatments for their patients, working with them to find phages that could be effective against the particular strain of bacteria infecting each patient. The newest paper, published today in the journal Clinical Infectious Diseases alongside collaborators from 20 institutions, dramatically scales up the body of published evidence on the effectiveness of the therapy.  “These are incredibly brave physicians, jumping off the ledge to do an experimental therapy to try to help patients who have no other options,” said Hatfull. “And each of these collaborations represents a marker that can move the field forward.” Looking at measures of patient health and whether samples from the patient still showed signs of Mycobacterium infections, the team found that the therapy was successful in 11 out of 20 cases. No patients showed any adverse reactions to the treatment. In another five patients the results of the therapy were inconclusive, and four patients showed no improvement. According to Hatfull, even these apparent failures are key to making the therapy available to more patients. “In some ways, those are the most interesting cases,” he said. “Understanding why they didn’t work is going to be important.” Several unexpected patterns emerged from the case studies. In 11 cases, researchers were unable to find more than one kind of phage that could kill the patient’s infection, even though standard practice would be to inject a cocktail of different viruses so the bacteria would be less likely to evolve resistance.


“If you’d asked me whether that was a good idea three years ago, I would have had a fit,” Hatfull said. “But we just didn’t observe resistance, and we didn’t see a failure of treatment from resistance even when using only a single phage.” In addition, the team saw that some patients’ immune systems attacked the viruses, but only in a few cases did their immune systems render the virus ineffective. And in some instances, the treatment was still successful despite such an immune reaction. The study paints an encouraging picture for the therapy, said Hatfull, and one that opens the possibility for new phage regimens that clinicians could use to maximize the treatment’s chance of success.  Along with the study’s significance to patients facing Mycobacterium infections, it also represents a substantial advance for the wider field of phage therapy. A concern in some corners is that researchers may be only publishing case studies in which phage therapy is successful. “A series of consecutive case studies, where we’re not cherry-picking, is a much more transparent way of looking to see what works and what doesn’t,” said Hatfull. “This adds considerable weight to the sense that the therapy is safe.” The lab continues to provide phages for more patients — while at the same time conducting research to widen the funnel that narrowed an initial group of 200 patients down to 20, in the hopes of providing treatment to more people with no other options. “We’ve not yet figured out how to find or engineer phages that will get every strain of bacteria in these patients,” said Hatfull. “That represents one of the major challenges ahead.”


Research cited published in Clinical Infectious Diseases (June 9, 2022): 

Via Juan Lama
No comment yet.
Scooped by Dr. Stefan Gruenwald!

Cough is mediated by a complex set of neurobiological processes involving the peripheral nervous system, brainstem and higher brain

Cough is mediated by a complex set of neurobiological processes involving the peripheral nervous system, brainstem and higher brain | Amazing Science |

Chronic cough is globally prevalent across all age groups. This disorder is challenging to treat because many pulmonary and extra-pulmonary conditions can present with chronic cough, and cough can also be present without any identifiable underlying cause or be refractory to therapies that improve associated conditions. Most patients with chronic cough have cough hypersensitivity, which is characterized by increased neural responsiveness to a range of stimuli that affect the airways and lungs, and other tissues innervated by common nerve supplies.


Cough hypersensitivity presents as excessive coughing often in response to relatively innocuous stimuli, causing significant psychophysical morbidity and affecting patients’ quality of life. Understanding the mechanism(s) that contribute to cough hypersensitivity and excessive coughing in different patient populations and across the lifespan is advancing and has contributed to the development of new therapies for chronic cough in adults. Owing to differences in the pathology, the organs involved and individual patient factors, treatment of chronic cough is progressing towards a personalized approach, and, in the future, novel ways to endotype patients with cough may prove valuable in management.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Minerva: Solving Math and Physics Problems with Language Models

Minerva: Solving Math and Physics Problems with Language Models | Amazing Science |

Language models have demonstrated remarkable performance on a variety of natural language tasks — indeed, a general lesson from many works, including BERTGPT-3Gopher, and PaLM, has been that neural networks trained on diverse data at large scale in an unsupervised way can perform well on a variety of tasks.

Quantitative reasoning is one area in which language models still fall far short of human-level performance.


Solving mathematical and scientific questions requires a combination of skills, including correctly parsing a question with natural language and mathematical notation, recalling relevant formulas and constants, and generating step-by-step solutions involving numerical calculations and symbolic manipulation. Due to these challenges, it is often believed that solving quantitative reasoning problems using machine learning will require significant advancements in model architecture and training techniques, granting models access to external tools such as Python interpreters, or possibly a more profound paradigm shift.


In “Solving Quantitative Reasoning Problems With Language Models” (to be released soon on the arXiv), we present Minerva, a language model capable of solving mathematical and scientific questions using step-by-step reasoning. We show that by focusing on collecting training data that is relevant for quantitative reasoning problems, training models at scale, and employing best-in-class inference techniques, we achieve significant performance gains on a variety of difficult quantitative reasoning tasks.


Minerva solves such problems by generating solutions that include numerical calculations and symbolic manipulation without relying on external tools such as a calculator. The model parses and answers mathematical questions using a mix of natural language and mathematical notation. Minerva combines several techniques, including few-shot promptingchain of thought or scratchpad prompting, and majority voting, to achieve state-of-the-art performance on STEM reasoning tasks. You can explore Minerva’s output with our interactive sample explorer!


A Model Built for Multi-step Quantitative Reasoning

To promote quantitative reasoning, Minerva builds on the Pathways Language Model (PaLM), with further training on a 118GB dataset of scientific papers from the arXiv preprint server and web pages that contain mathematical expressions using LaTeXMathJax, or other mathematical typesetting formats. Standard text cleaning procedures often remove symbols and formatting that are essential to the semantic meaning of mathematical expressions. By maintaining this information in the training data, the model learns to converse using standard mathematical notation.
Minerva also incorporates recent prompting and evaluation techniques to better solve mathematical questions. These include chain of thought or scratchpad prompting — where Minerva is prompted with several step-by-step solutions to existing questions before being presented with a new question — and majority voting. Like most language models, Minerva assigns probabilities to different possible outputs. When answering a question, rather than taking the single solution Minerva scores as most likely, multiple solutions are generated by sampling stochastically from all possible outputs. These solutions are different (e.g., the steps are not identical), but often arrive at the same final answer. Minerva uses majority voting on these sampled solutions, taking the most common result as the conclusive final answer.
No comment yet.
Scooped by Dr. Stefan Gruenwald!

Grey wolf genomic history reveals a dual ancestry of dogs

Grey wolf genomic history reveals a dual ancestry of dogs | Amazing Science |

The grey wolf (Canis lupus) has been present across most of the northern hemisphere for the last few hundred thousand years and, unlike many other large mammals, did not go extinct in the Late Pleistocene. Studies of present-day genomes have found that current population structure formed mostly in the last ~30,000–20,000 years9,10,11, or roughly since the Last Glacial Maximum (LGM; ~28–23 thousand years ago (ka)12). Siberian wolves predating the LGM have ancestries that are largely basal to present-day diversity, which has led to suggestions that many pre-LGM wolf lineages went extinct13,14. Among the central questions is thus to what extent the global wolf population was subject to extinction processes or responded to climate change with new adaptations.


While it is clear that grey wolves gave rise to dogs, there is no consensus regarding when, where and how this happened1,2,3,4,5,6,7,8. Skeletal remains attributable to the present-day dog lineage appear archaeologically by 14 ka15, and genetic estimates of when the ancestors of dogs and modern wolves diverged range from 40–14 ka9,13,16. However, genetic data from modern and ancient dogs coupled with modern wolves, to which previous studies were largely restricted, may not be able to resolve the origin of dogs. Genetic diversity within dogs is affected by their dynamic history and is unable to confidently pinpoint an origin.


Relationships to modern wolves can likewise be affected by local extinction and gene flow since domestication6,9. Regions where early dogs have been found do not necessarily imply places of origin either, as the existence of earlier dogs elsewhere cannot be excluded. Instead, the origin of dogs could be resolved if wolf genetic diversity across space and time was exhaustively characterized and it could be determined which populations were closest to the ancestors of dogs.


A research team now sequenced 66 new ancient wolf genomes from Europe, Siberia and north-western North America to a median of 1× coverage (range, 0.02–13×) and incorporated five previously sequenced ancient wolf genomes14,17 and increased coverage for one13. They also sequenced an ancient dhole genome from the Caucasus, contextually dated to >70 ka, to serve as an outgroup. Fractions of X-chromosome DNA showed that 69% of the wolves were male (95% confidence interval (CI), 57–80%; P = 0.0013, binomial test), mirroring male over-representation among ancient genomes from woolly mammoths18, bison19, brown bears19 and domestic dogs8. The total dataset spans the last 100,000 years.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Novel Biofinder instrument could advance detection of extraterrestrial life

Novel Biofinder instrument could advance detection of extraterrestrial life | Amazing Science |

An innovative scientific instrument, the Compact Color Biofinder, developed by a team of University of Hawai‘i at Mānoa researchers, may change the game in the search for signs of extraterrestrial life.


Most biological materials, for example, amino acids, fossils, sedimentary rocks, plants, microbes, proteins and lipids, have strong organic fluorescence signals that can be detected by specialized scanning cameras. In a study published in Nature Scientific Reports recently, the research team reported that the Biofinder is so sensitive that it can accurately detect the bio-residue in fish fossils from the 34-56 million year-old Green River formation.


"The Biofinder is the first system of its kind," said Anupam Misra, lead instrument developer and researcher at the Hawai'i Institute of Geophysics and Planetology at the UH Manoa School of Ocean and Earth Science and Technology (SOEST). "At present, there is no other equipment that can detect minute amounts of bio-residue on a rock during the daytime. Additional strengths of the Biofinder are that it works from a distance of several meters, takes video and can quickly scan a large area."


Though the Biofinder was first developed in 2012 by Misra, advances supported by the NASA PICASSO program culminated in the latest color version of the compact Biofinder. Finding evidence of biological residue in a vast planetary landscape is an enormous challenge. So, the team tested the Biofinder's detection abilities on the ancient Green River fish fossils and corroborated the results through laboratory spectroscopy analysis, scanning electron microscopy and fluorescence lifetime imaging microscopy.


"There are some unknowns regarding how quickly bio-residues are replaced by minerals in the fossilization process," said Misra. "However, our findings confirm once more that biological residues can survive millions of years, and that using biofluorescence imaging effectively detects these trace residues in real time."


The search for life -- which may be existing or extinct -- on planetary bodies is one of the major goals of planetary exploration missions conducted by NASA and other international space agencies. "If the Biofinder were mounted on a rover on Mars or another planet, we would be able to rapidly scan large areas quickly to detect evidence of past life, even if the organism was small, not easy to see with our eyes, and dead for many millions of years," said Misra. "We anticipate that fluorescence imaging will be critical in future NASA missions to detect organics and the existence of life on other planetary bodies."


"The Biofinder's capabilities would be critical for NASA's Planetary Protection program, for the accurate and no-invasive detection of contaminants such as microbes or extraterrestrial biohazards to or from planet Earth," said Sonia J. Rowley, the team biologist and co-author on the study. Misra and colleagues are applying to have the opportunity to send the Biofinder on a future NASA mission. "The detection of such biomarkers would constitute groundbreaking evidence for life outside of planet Earth," said Misra.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Gaia continues its quest for the ultimate sky map and has already analyzed the positions of 2 billion objects

Gaia continues its quest for the ultimate sky map and has already analyzed the positions of 2 billion objects | Amazing Science |
Europe's Gaia space camera tracks everything that shines or moves in our Milky Way galaxy.


Europe's Gaia telescope has dropped its latest batch of data as it seeks to assemble the largest catalog of light sources in the sky. 


It is becoming a discovery machine like no other. Stars, asteroids and distant, bright galaxies - anything that can be visibly pinpointed is having its vital statistics measured by the observatory. Gaia has already mapped the positions of nearly two billion objects. Now, it can reveal more about their make-up.

"Essentially, previously, we could say very precisely where they are; now we can say what they are," Prof Nick Walton, from Cambridge University and a member of the Gaia science team, told BBC News. The European Space Agency's (Esa) Gaia satellite was launched in 2013 and placed a million miles from Earth.


It looks a bit like a spinning top hat. And as it rotates, the telescope uses its British-built billion-pixel camera to track everything that shines or moves - with astonishing accuracy. This is especially important when trying to measure distances to objects, which Gaia achieves by tracking how these targets wobble ever so slightly on the sky as it circles the Sun - a neat form of trigonometry that has now been practiced on 1.8 billion stars in, or very near, our Milky Way galaxy.

  • As the Earth goes around the Sun, relatively nearby stars appear to move against the "fixed" stars that are even further away
  • Because we know the Sun-Earth distance, we can use the parallax angle to work out the distance to the target star
  • But such angles are very small - less than one arcsecond for the nearest stars, or 0.05% of the full Moon's diameter
  • Gaia is making repeat observations to reduce measurement errors down to seven micro-arcseconds for the very brightest stars
  • Parallaxes are used to anchor other, more indirect techniques on the 'ladder' deployed to measure the most far-flung distances

In the previous release of data, in December 2020, Gaia also revealed basic brightness and color information on these stars.

The new data dump reveals spectroscopy information as well.

Spectroscopy slices the light coming from stars into its constituent colors, to reveal the chemistry, temperature, mass, age and velocity of the targets under study.

And for an important subset of stars - some 33 million - it has allowed Gaia scientists to determine how quickly these objects are moving towards or away from Earth.


Combined with their previously established movement across the sky, this means we now have their full three-dimensional behavior. Such information will give researchers even keener insights on how the Milky Way galaxy is structured and is evolving - from the past, into the future.


Gaia's data haul now includes:

  • two billion light sources - mostly stars but also many Solar System objects and some beyond the Milky Way
  • spectroscopic detail revealing temperature, chemistry, mass and age for 100s of millions of objects
  • 1.9 million quasars - distant galaxies where a voracious central black hole is powering light emission
  • 156,000 asteroids - critical for understanding their origin and possibility of them passing close to Earth
No comment yet.
Scooped by Dr. Stefan Gruenwald!

How AI creates photorealistic images from simple text

How AI creates photorealistic images from simple text | Amazing Science |

Have you ever seen a puppy in a nest emerging from a cracked egg? What about a photo that’s overlooking a steampunk city with airships? Or a picture of two robots having a romantic evening at the movies? These might sound far-fetched, but a novel type of machine learning technology called text-to-image generation makes them possible. These models can generate high-quality, photorealistic images from a simple text prompt.


Within Google Research, our scientists and engineers have been exploring text-to-image generation using a variety of AI techniques. After a lot of testing we recently announced two new text-to-image models — Imagen and Parti. Both have the ability to generate photorealistic images but use different approaches. We want to share a little more about how these models work and their potential.

How text-to-image models work

With text-to-image models, people provide a text description and the models produce images matching the description as closely as possible. This can be something as simple as “an apple” or “a cat sitting on a couch” to more complex details, interactions and descriptive indicators like “a cute sloth holding a small treasure chest. A bright golden glow is coming from the chest. In the past few years, ML models have been trained on large image datasets with corresponding textual descriptions, resulting in higher quality images and a broader range of descriptions. This has sparked major breakthroughs in this area, including Open AI’s DALL-E 2.

How Imagen and Parti work

Imagen and Parti build on previous models. Transformer models are able to process words in relationship to one another in a sentence. They are foundational to how we represent text in our text-to-image models. Both models also use a new technique that helps generate images that more closely match the text description. While Imagen and Parti use similar technology, they pursue different, but complementary strategies.


Imagen is a Diffusion model, which learns to convert a pattern of random dots to images. These images first start as low resolution and then progressively increase in resolution. Recently, Diffusion models have seen success in both image and audio tasks like enhancing image resolution, recoloring black and white photos, editing regions of an image, uncropping images, and text-to-speech synthesis.


Parti’s approach first converts a collection of images into a sequence of code entries, similar to puzzle pieces. A given text prompt is then translated into these code entries and a new image is created. This approach takes advantage of existing research and infrastructure for large language models such as PaLM and is critical for handling long, complex text prompts and producing high-quality images.


These models have many limitations. For example, neither can reliably produce specific counts of objects (e.g. “ten apples”), nor place them correctly based on specific spatial descriptions (e.g. “a red sphere to the left of a blue block with a yellow triangle on it”). Also, as prompts become more complex, the models begin to falter, either missing details or introducing details that were not provided in the prompt. These behaviors are a result of several shortcomings, including lack of explicit training material, limited data representation, and lack of 3D awareness. We hope to address these gaps through broader representations and more effective integration into the text-to-image generation process.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Exercise benefits from a pill? Science is closer to that goal

Exercise benefits from a pill? Science is closer to that goal | Amazing Science |

Researchers at Baylor College of Medicine, Stanford School of Medicine and collaborating institutions report today in the journal Nature that they have identified a molecule in the blood that is produced during exercise and can effectively reduce food intake and obesity in mice. The findings improve our understanding of the physiological processes that underlie the interplay between exercise and hunger.


"Regular exercise has been proven to help weight loss, regulate appetite and improve the metabolic profile, especially for people who are overweight and obese," said co-corresponding author Dr. Yong Xu, professor of pediatrics- nutrition and molecular and cellular biology at Baylor. "If we can understand the mechanism by which exercise triggers these benefits, then we are closer to helping many people improve their health."


"We wanted to understand how exercise works at the molecular level to be able to capture some of its benefits," said co-corresponding author Jonathan Long, MD, assistant professor of pathology at Stanford Medicine and an Institute Scholar of Stanford ChEM-H (Chemistry, Engineering & Medicine for Human Health). "For example, older or frail people who cannot exercise enough, may one day benefit from taking a medication that can help slow down osteoporosis, heart disease or other conditions."


Xu, Long and their colleagues conducted comprehensive analyses of blood plasma compounds from mice following intense treadmill running. The most significantly induced molecule was a modified amino acid called Lac-Phe. It is synthesized from lactate (a byproduct of strenuous exercise that is responsible for the burning sensation in muscles) and phenylalanine, an amino acid that is one of the regular building blocks of proteins.


In mice with diet-induced obesity (fed a high-fat diet), a high dose of Lac-Phe suppressed food intake by about 50% compared to control mice over a period of 12 hours without affecting their movement or energy expenditure. When administered to the mice for 10 days, Lac-Phe reduced cumulative food intake and body weight (owing to loss of body fat) and improved glucose tolerance.

The researchers also identified an enzyme called CNDP2 that is involved in the production of Lac-Phe and showed that mice lacking this enzyme did not lose as much weight on an exercise regime as a control group on the same exercise plan.


Interestingly, the team also found robust elevations in plasma Lac-Phe levels following physical activity in racehorses and humans. Data from a human exercise cohort showed that sprint exercise induced the most dramatic increase in plasma Lac-Phe, followed by resistance training and then endurance training. "This suggests that Lac-Phe is an ancient and conserved system that regulates feeding and is associated with physical activity in many animal species," Long said.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

The Gaia space telescope rocks the science of asteroids

The Gaia space telescope rocks the science of asteroids | Amazing Science |

The Gaia space mission of the European Space Agency ESA is constructing an ultra-precise three-dimensional map of our Milky Way galaxy, observing almost two billion stars or roughly one percent of all the stars in our galaxy. Gaia was launched in December 2013 and has collected science data from July 2014.


Recenlty, ESA released Gaia data in Data Release 3 (DR3). Gaia data allows for the derivation of asteroid and exoplanet orbits and physical properties. The data helps unveil the origin and future evolution of the Solar System and the Milky Way and helps understand stellar and planetary-system evolution and our place in the cosmos.


Gaia revolves about its axis slowly in about six hours and is composed of two optical space telescopes. Three science instruments allow for accurate determination of stellar positions and velocities as well as the spectral properties. Gaia resides at about 1,5 million kilometers from the Earth in the anti-Sun direction, where it orbits the Sun together with the Earth in the proximity of the so-called Sun-Earth Lagrange L2-point.


Gaia DR3 on June 13, 2022 was significant across astronomy. Some 50 scientific articles are being published with DR3, of which nine articles have been devoted to underscoring the exceptionally significant potential of DR3 for future research. The new DR3 data comprises, for example, the chemical compositions, temperatures, colors, masses, brightnesses, ages, and radial velocities of stars. DR3 includes the largest ever binary star catalog for the Milky Way, more than 150 000 Solar System objects, largely asteroids but also planetary satellites, as well as millions of galaxies and quasars beyond the Milky Way.


There are so many revolutionary advances that it is difficult to pinpoint a single most significant advance. Based on Gaia DR3, Finnish researchers will change the conception of asteroids in our Solar System, exoplanets and stars in our Milky Way galaxy, as well as galaxies themselves, including the Milky Way and its surrounding satellite galaxies. Returning to our home planet, Gaia will produce an ultraprecise reference frame for navigation and positioning, says Academy Professor Karri Muinonen from the University of Helsinki.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

The dark secret behind those cute AI-generated animal images

The dark secret behind those cute AI-generated animal images | Amazing Science |

Another month, another flood of weird and wonderful images generated by an artificial intelligence. In April 2022, OpenAI showed off its new picture-making neural network, DALL-E 2, which could produce remarkable high-res images of almost anything it was asked to. It outstripped the original DALL-E in almost every way.


Now, just a few weeks later, Google Brain has revealed its own image-making AI, called Imagen. And it performs even better than DALL-E 2: it scores higher on a standard measure for rating the quality of computer-generated images, and the pictures it produced were preferred by a group of human judges.

“We’re living through the AI space race!” one Twitter user commented. “The stock image industry is officially toast,” tweeted another.


Many of Imagen’s images are indeed jaw-dropping. At a glance, some of its outdoor scenes could have been lifted from the pages of National Geographic. Marketing teams could use Imagen to produce billboard-ready advertisements with just a few clicks.


But as OpenAI did with DALL-E, Google is going all in on cuteness. Both firms promote their tools with pictures of anthropomorphic animals doing adorable things: a fuzzy panda dressed as a chef making dough, a corgi sitting in a house made of sushi, a teddy bear swimming the 400-meter butterfly at the Olympics—and it goes on.


There’s a technical, as well as PR, reason for this. Mixing concepts like “fuzzy panda” and “making dough” forces the neural network to learn how to manipulate those concepts in a way that makes sense. But the cuteness hides a darker side to these tools, one that the public doesn’t get to see because it would reveal the ugly truth about how they are created.


Most of the images that OpenAI and Google make public are cherry-picked. We only see cute images that match their prompts with uncanny accuracy—that’s to be expected. But we also see no images that contain hateful stereotypes, racism, or misogyny. There is no violent, sexist imagery. There is no panda porn. And from what we know about how these tools are built—there should be. So, what's going on?


For now, the solution is to keep them caged up. OpenAI is making DALL-E 2 available only to a handful of trusted users; Google has no plans to release Imagen. That’s fine if these were simply proprietary tools. But these firms are pushing the boundaries of what AI can do and their work shapes the kind of AI that all of us live with. They are creating new marvels, but also new horrors— and moving on with a shrug. When Google’s in-house ethics team raised problems with the large language models, in 2020 it sparked a fight that ended with two of its leading researchers being fired.


Large language models and image-making AIs have the potential to be world-changing technologies, but only if their toxicity is tamed. This will require a lot more research. There are small steps to open these kinds of neural network up for widespread study. A few weeks ago Meta released a large language model to researchers, warts and all. And Hugging Face is set to release its open-source version of GPT-3 in the next couple of months. 


For now, enjoy the cute teddies.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

'Fantastic giant tortoise,' believed extinct, confirmed alive in the Galápagos after more than a century

'Fantastic giant tortoise,' believed extinct, confirmed alive in the Galápagos after more than a century | Amazing Science |

A tortoise from a Galápagos species long believed extinct has been found alive. The tortoise, named Fernanda after her Fernandina Island home, is the first of her species identified in more than a century.


The Fernandina Island Galápagos giant tortoise (Chelonoidis phantasticus, or "fantastic giant tortoise") was known only from a single specimen, collected in 1906. The discovery in 2019 of a female tortoise living on Fernandina Island provided the opportunity to determine if the species lives on. By sequencing the genomes of both the living individual and the museum specimen, and comparing them to the other 13 species of Galápagos giant tortoises, Princeton's Stephen Gaughran showed that the two known Fernandina tortoises are members of the same species, genetically distinct from all others. He is co-first author on a paper in the current issue of Communications Biology confirming her species' continued existence.


"For many years it was thought that the original specimen collected in 1906 had been transplanted to the island, as it was the only one of its kind," said Peter Grant, Princeton's Class of 1877 Professor of Zoology, Emeritus and an emeritus professor of ecology and evolutionary biology who has spent more than 40 years studying evolution in the Galápagos islands. "It now seems to be one of a very few that were alive a century ago."


When Fernanda was discovered, many ecologists doubted that she was actually a native phantasticus tortoise. She lacks the striking saddleback flaring of the male historical specimen, though scientists speculated that her obviously stunted growth may have distorted her features. Tortoises can't swim from one island to another, but they do float, and they can be carried from one Galápagos island to another during hurricanes or other major storms. There are also historical records of seafarers moving the tortoises between islands.


"Like many people, my initial suspicion was that this was not a native tortoise of Fernandina Island," said Gaughran, a postdoctoral research fellow in ecology and evolutionary biology at Princeton.


To determine Fernanda's species definitively, Gaughran sequenced her complete genome and compared it to the genome he was able to recover from the specimen collected in 1906. He also compared those two genomes to samples from the other 13 species of Galápagos tortoises -- three individuals from each of the 12 living species, and one individual of the extinct C. abingdonii.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

New evidence emerges about when, where, and how chickens were domesticated

New evidence emerges about when, where, and how chickens were domesticated | Amazing Science |

New research transforms our understanding of the circumstances and timing of the domestication of chickens, their spread across Asia into the west, and reveals the changing way in which they were perceived in societies over the past 3,500 years.


Experts have found that an association with rice farming likely started a process that has led to chickens becoming one of the world's most numerous animals. They have also found evidence that chickens were initially regarded as exotica, and only several centuries later used as a source of 'food'.


Previous efforts have claimed that chickens were domesticated up to 10,000 years ago in China, Southeast Asia, or India, and that chickens were present in Europe over 7,000 years ago. The new studies show this is wrong, and that the driving force behind chicken domestication was the arrival of dry rice farming into southeast Asia where their wild ancestor, the red jungle fowl, lived. Dry rice farming acted as a magnet drawing wild jungle fowl down from the trees, and kickstarting a closer relationship between people and the jungle fowl that resulted in chickens.


This domestication process was underway by around 1,500 BC in the Southeast Asia peninsula. The research suggests that chickens were then transported first across Asia and then throughout the Mediterranean along routes used by early Greek, Etruscan and Phoenician maritime traders.


During the Iron Age in Europe, chickens were venerated and generally not regarded as food. The studies have shown that several of the earliest chickens are buried alone and un-butchered, and many are also found buried with people. Males were often buried with cockerels and females with hens. The Roman Empire then helped to popularise chickens and eggs as food. For example, in Britain, chickens were not regularly consumed until the third century AD, mostly in urban and military sites.


The international research team re-evaluated chicken remains found in more than 600 sites in 89 countries. They examined the skeletons, burial location and historical records regarding the societies and cultures where the bones were found. The oldest bones of a definite domestic chicken were found at Neolithic Ban Non Wat in central Thailand, and date to between 1,650 and 1,250 BC.


The team of experts also used radiocarbon dating to establish the age of 23 of the proposed earliest chickens found in western Eurasia and north-west Africa. Most of the bones were far more recent than previously thought. The results dispel claims of chickens in Europe before the first millennium BC and indicate that they did not arrive until around 800 BC. Then, after arriving in the Mediterranean region, it took almost 1,000 years longer for chickens to become established in the colder climates of Scotland, Ireland, Scandinavia and Iceland.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

Ground zero for the Black Death finally found after 600 years

Ground zero for the Black Death finally found after 600 years | Amazing Science |

The origins of the deadly Black Death have been discovered more than 600 years after it entered the human population, scientists have said. The medieval, bubonic plague was first recorded in the 14th century and was the start of a near 500-year-long wave of killer diseases termed the Second Plague Pandemic. The Black Death killed millions and was considered one of the largest infectious disease catastrophes in human history. Despite years of research, the geographic and chronological origin of the disease remained a mystery. But now researchers believe the Black Death first originated in North Kyrgyzstan in the late 1330s. The team, from Scotland’s University of Stirling and Germany’s Max Planck Institute and University of Tubingen, analysed ancient DNA (aDNA) taken from the teeth of skeletons discovered in cemeteries near Lake Issyk Kul in the Tian Shan region of Kyrgyzstan.

They were drawn to these sites after identifying a huge spike in the number of burials there in 1338 and 1339, according to University of Stirling historian Dr Philip Slavin, who helped make the discovery. The team found the cemeteries, at Kara-Djigach and Burana, had already been excavated in the late 1880s, with about 30 skeletons taken from the graves, but were able to trace them and analyse DNA taken from the teeth of seven individuals.


The sequencing, which determines the DNA structure, showed three individuals carried Yersinia pestis, a bacterium which is linked to the beginning of the Black Death outbreak before it arrived in Europe. ‘Our study puts to rest one of the biggest and most fascinating questions in history and determines when and where the single most notorious and infamous killer of humans began,’ Dr Slavin said.

No comment yet.
Scooped by Dr. Stefan Gruenwald!

NASA citizen scientists turn to virtual reality to explore our galaxy in a new way and present their first discoveries

NASA citizen scientists turn to virtual reality to explore our galaxy in a new way and present their first discoveries | Amazing Science |

At the 240th meeting of the American Astronomical Society (AAS) in Pasadena, California, NASA's Disk Detective citizen scientists presented their first virtual reality-enabled discoveries, including the ages of 10 stars with dusty disks, on Tuesday, June 14, 2022.


NASA's Disk Detective program allows volunteers to collaborate with professional scientists to help search for dusty disks around nearby stars, revealing clues to the early lives of stars and the ingredients of planets.


A team led by Thomas Grubb at NASA's Goddard Space Flight Center in Greenbelt, Maryland, developed a custom virtual reality program, called PointCloudsVR, for Disk Detective to help scientists explore the galaxy in a brand new way.


Susan Higashio, lead author of the Disk Detective team's new paper, and colleagues found 10 of the citizen-science star discoveries in young stellar associations, allowing them to determine the ages for each dusty disk star, ranging from about 18 to 133 million years old. One of these stars has proven to be an oddball because it seems to have too much dust for its age of 45 million years.


They also identified a young, previously unknown stellar association - the team informally calls this group "Smethells 165" after the name of its brightest star, cataloged by astronomer William George Smethells. "Young stars like the ones we study often form in groups – and when we view our data in VR it enables us to see things from a new perspective. It can be easier to spot these groups," said Susan Higashio, who flew through the Milky Way with a VR headset strapped to her face.

No comment yet.