A few Elderly Have Memories of 25 Year Olds

For the first time, researchers have used fMRI to understand how some older adults can learn and remember new information as well as a 25-year-old.

From:  MASSACHUSETTS GENERAL HOSPITAL

July 6, 2021 -- BOSTON -- As we age, our brains typically undergo a slow process of atrophy, causing less robust communication between various brain regions, which leads to declining memory and other cognitive functions. But a rare group of older individuals called "superagers" have been shown to learn and recall novel information as well as a 25-year-old. Investigators from Massachusetts General Hospital (MGH) have now identified the brain activity that underlies superagers' superior memory. "This is the first time we have images of the function of superagers' brains as they actively learn and remember new information," says Alexandra Touroutoglou, PhD, director of Imaging Operations at MGH's Frontotemporal Disorders Unit and senior author of the paper published in Cerebral Cortex.

In 2016, Touroutoglou and her fellow researchers identified a group of adults older than 65 with remarkable performance on memory tests. The superagers are participants in an ongoing longitudinal study of aging at MGH led by Bradford Dickerson, MD, director of the Frontotemporal Disorders Unit at MGH, and Lisa Feldman Barrett, PhD, a research scientist in Psychiatry at MGH. "Using MRI, we found that the structure of superagers' brains and the connectivity of their neural networks more closely resemble the brains of young adults; superagers had avoided the brain atrophy typically seen in older adults," says Touroutoglou.

In the new study, the investigators gave 40 adults with a mean age of 67 a very challenging memory test while their brains were imaged using functional magnetic resonance imaging (fMRI), which, unlike typical MRI, shows the activity of different brain areas during tasks. Forty-one young adults (mean age of 25) also took the same memory test while their brains were imaged. The participants first viewed 80 pictures of faces or scenes that were each paired with an adjective, such as a cityscape paired with the word "industrial" or a male face paired with the word "average." Their first task was to determine whether the word matched the image, a process called encoding. After 10 minutes, participants were presented with the 80 image-word pairs they had just learned, an additional 40 pairs of new words and images, and 40 rearranged pairs consisting of words and images they had previously seen. Their second task was to recall whether they had previously seen each specific word-picture pair, or whether they were looking at a new or rearranged pair.

While the participants were in the scanner, the researchers paid close attention to the visual cortex, which is the area of the brain that processes what you see and is particularly sensitive to aging. "In the visual cortex, there are populations of neurons that are selectively involved in processing different categories of images, such as faces, houses or scenes," says lead author Yuta Katsumi, PhD, a postdoctoral fellow in Psychiatry at MGH. "This selective function of each group of neurons makes them more efficient at processing what you see and creating a distinct memory of those images, which can then easily be retrieved."

During aging, this selectivity, called neural differentiation, diminishes and the group of neurons that once responded primarily to faces now activates for other images. The brain now has difficulty creating unique neural activation patterns for different types of images, which means it is making less distinctive mental representations of what the person is seeing. That's one reason older individuals have trouble remembering when they may have seen a television show, read an article, or eaten a specific meal.

But in the fMRI study, the superagers' memory performance was indistinguishable from the 25-year-olds', and their brains' visual cortex maintained youthful activity patterns. "The superagers had maintained the same high level of neural differentiation, or selectivity, as a young adult," says Katsumi. "Their brains enabled them to create distinct representations of the different categories of visual information so that they could accurately remember the image-word pairs."

An important question that researchers still must answer is whether "superagers' brains were always more efficient than their peers, or whether, over time, they developed mechanisms to compensate for the decline of the aging brain," says Touroutoglou.

Previous studies have shown that training can increase the selectivity of brain regions, which may be a potential intervention to delay or prevent the decline in neural differentiation in normal aging adults and make their brains more like those of superagers. Currently the researchers are conducting a clinical trial to evaluate whether noninvasive electromagnetic stimulation, which delivers an electrical current to targeted areas of the brain, can improve memory in older adults. The researchers also plan to study different brain regions to further understand how superagers learn and remember, and they will examine lifestyle and other factors that might contribute to superagers' amazing memory.

          https://www.eurekalert.org/pub_releases/2021-07/mgh-srs070621.php

We Must Celebrate USA's Natural Gas Boom

Natural Gas Is the Greenest Energy Currently Available “Off the Shelf” at Present Level of Efficient Technologies

By David Callahan

July 05, 2021 -- As we celebrate the 4^th of July holiday, America’s national security should be front of mind. As such, the energy freedoms that are made possible by domestic natural gas development deserve to be recognized and celebrated this Independence Day weekend.

Energy security has been a national priority for generations, and was brought to the forefront of policy discussions during the energy crises of the 1970s. Two years ago, thanks to the continued

success of shale development, America achieved that elusive national security goal when the U.S. became a net energy exporter for the first time since the mid-1950s.

It’s an achievement celebrated broadly, with Energy Secretary Jennifer Granholm telling members of Congress during her confirmation hearing that “hydraulic fracturing and horizontal drilling…have certainly contributed to the nation’s energy security.”

The widespread adoption of these American-born technologies has enabled the U.S. to become the world’s largest natural gas and oil producer while also reducing energy sector emissions faster than any other nation.

Compared to peak 2005 levels, Pennsylvania power sector CO2 emissions have declined 41 percent, and state data shows sulfur oxides and nitrogen oxide levels plummeted 93 percent and 81 percent, respectively, as more natural gas-fired generation came online during that time frame. This leads to significantly fewer respiratory illnesses, which often disproportionately affect children and senior citizens.

While these energy development successes are being realized in the U.S., energy poverty still burdens many countries across the world. America has the opportunity – and responsibility – to take a leadership role in reducing global energy poverty and help less fortunate nations find a new pathway towards energy abundance and economic growth while also reducing emissions.

Clean, American-produced natural gas is capable of meeting the ever-growing global energy demand, with the natural gas produced here being done so under the most rigorous environmental standards. Furthermore, according to the International Energy Agency’s methane tracker, the U.S. has among the lowest methane emission intensity of all natural gas and oil producing nations.

Domestically, methane intensity associated with Appalachian production is by far the lowest of the top nine hydrocarbon-producing basins in the U.S., a joint report by the Clean Air Task Force and Ceres found.

This environmental integrity holds true among other air emissions, such as CO2, where Appalachia is also associated with the lowest emission intensity, according to a recent Rystad Energy analysis, bringing the basin “to the top quartile among all oil and gas fields globally,” senior Rystad analyst Emily McClain said.

With the overwhelming environmental benefits associated with U.S.-produced energy, it’s no surprise the global demand for American natural gas reached record levels last year and is on track to keep growing as new technology and efficiency practices come underway.

But using our domestic resources for global climate progress hinges on our ability to build new energy infrastructure and pass inclusive energy policies that recognize the many benefits of this critical resource for the environment and the economy.

Policies that ignore natural gas’ long and short-term benefits threaten the energy security America worked so hard to achieve. Moreover, impeding domestic energy production would eliminate all means of economic and energy security, while forcing consumers to rely on energy imports from countries who lack the same industry-leading environmental oversight and regulatory framework.

As we celebrate Independence Day, let’s not forget that energy security is national security, driven by the men and women of the natural gas industry innovating and working in overdrive to meet clean energy and economic development goals worldwide.

[David Callahan is president of the Pittsburgh-based Marcellus Shale Coalition. To learn more, visit marcelluscoalition.org.]

https://www.realclearenergy.org/articles/2021/07/05/why_we_must_celebrate_americas_natural_gas_boom_784193.html

Instant Water Cleaning Method Considered “Millions of Times” Better than a Commercial Approach

Creation of hydrogen peroxide in situ could provide clean, drinkable water to communities in the poorest nations around the world

From: Cardiff University

July 1, 2021 -- A water disinfectant created on the spot using just hydrogen and the air around us is millions of times more effective at killing viruses and bacteria than traditional commercial methods, according to scientists from Cardiff University.

Reporting their findings today in the journal Nature Catalysis, the team say the results could revolutionize water disinfection technologies and present an unprecedented opportunity to provide clean water to communities that need it most.

Their new method works by using a catalyst made from gold and palladium that takes in hydrogen and oxygen to form hydrogen peroxide -- a commonly used disinfectant that is currently produced on an industrial scale.

Over four million tonnes of hydrogen peroxide are made in factories each year, where it is then transported to the places it is used and stored. This means that stabilizing chemicals are often added to the solutions during the production process to stop it degrading but these reduce its effectiveness as a disinfectant.

Another common approach to disinfecting water is the addition of chlorine; however, it has been shown that chlorine can react with naturally occurring compounds in water to form compounds which, in high doses, can be toxic to humans.

The ability to be able to produce hydrogen peroxide at the point of use would overcome both efficacy and safety issues currently associated with commercial methods.

In their study, the team tested the disinfection efficacy of commercially available hydrogen peroxide and chlorine compared to their new catalytic method.

Each was tested for its ability to kill Escherichia coli in identical conditions, followed by subsequent analysis to determine the processes by which the bacteria were killed using each method.

The team showed that as the catalyst brought the hydrogen and oxygen together to form hydrogen peroxide, it simultaneously produced a number of highly reactive compounds, known as reactive oxygen species (ROS), which the team demonstrated were responsible for the antibacterial and antiviral effect, and not the hydrogen peroxide itself.

The catalyst-based method was shown to be 10,000,000 times more potent at killing the bacteria than an equivalent amount of the industrial hydrogen peroxide, and over 100,000,000 times more effective than chlorination, under equivalent conditions.

In addition to this, the catalyst-based method was shown to be more effective at killing the bacteria and viruses in a shorter space of time compared to the other two compounds.

It is estimated that around 785 million people lack access to water and 2.7 billion experience water scarcity at least one month a year.

In addition to this, inadequate sanitation -- a problem for around 2.4 billion people around the world -- can lead to deadly diarrheal diseases, including cholera and typhoid fever, and other water-borne illnesses.

Co-author of the study Professor Graham Hutchings, Regius Professor of Chemistry at the Cardiff Catalysis Institute, said: "The significantly enhanced bactericidal and virucidal activities achieved when reacting hydrogen and oxygen using our catalyst, rather than using commercial hydrogen peroxide or chlorination shows the potential for revolutionizing water disinfection technologies around the world.

"We now have proven one-step process where, besides the catalyst, inputs of contaminated water and electricity are the only requirements to attain disinfection.

"Crucially, this process presents the opportunity to rapidly disinfect water over timescales in which conventional methods are ineffective, whilst also preventing the formation of hazardous compounds and biofilms, which can help bacteria and viruses to thrive."

               https://www.sciencedaily.com/releases/2021/07/210701112652.htm

Tax Avoidance and Tax Evasion

Tax noncompliance (informally tax avoision) is a range of activities that are unfavorable to a government's tax system. This may include tax avoidance, which is tax reduction by legal means, and tax evasion which is the criminal non-payment of tax liabilities.  The use of the term 'noncompliance' is used differently by different authors.  Its most general use describes non-compliant behaviors with respect to different institutional rules resulting in what Edgar L. Feige calls unobserved economies.  Non-compliance with fiscal rules of taxation gives rise to unreported income and a tax gap that Feige estimates to be in the neighborhood of $500 billion annually for the United States.

In the United States, the use of the term 'noncompliance' often refers only to illegal misreporting.  Laws known as a General Anti-Avoidance Rule (GAAR) statutes which prohibit "tax aggressive" avoidance have been passed in several developed countries including the United States (since 2010), Canada, Australia, New Zealand, South Africa, Norway and Hong Kong.  In addition, judicial doctrines have accomplished the similar purpose, notably in the United States through the "business purpose" and "economic substance" doctrines established in Gregory v. Helvering. Though the specifics may vary according to jurisdiction, these rules invalidate tax avoidance which is technically legal but not for a business purpose or in violation of the spirit of the tax code.  Related terms for tax avoidance include tax planning and tax sheltering.

Individuals that do not comply with tax payment include tax protesters and tax resisters. Tax protesters attempt to evade the payment of taxes using alternative interpretations of the tax law, while tax resisters refuse to pay a tax for conscientious reasons. In the United States, tax protesters believe that taxation under the Federal Reserve is unconstitutional, while tax resisters are more concerned with not paying for particular government policies that they oppose. Because taxation is often perceived as onerous, governments have struggled with tax noncompliance since the earliest of times.

Differences between Avoidance and Evasion

The use of the terms tax avoidance and tax evasion can vary depending on the jurisdiction. In general, the term "evasion" applies to illegal actions and "avoidance" to actions within the law. The term "mitigation" is also used in some jurisdictions to further distinguish actions within the original purpose of the relevant provision from those actions that are within the letter of the law, but do not achieve its purpose.

As the difference between the two concepts is currently becoming less clear, law professor Allison Christians deplores the condition that morality is being cited as a criterion instead of the rule of law.

History

An avoidance/evasion distinction along the lines of the present distinction has long been recognised but at first there was no terminology to express it. In 1860 Turner LJ suggested evasion/contravention (where evasion stood for the lawful side of the divide): Fisher v Brierly.  In 1900 the distinction was noted as two meanings of the word "evade": Bullivant v AG.  The technical use of the words avoidance/evasion in the modern sense originated in the US where it was well established by the 1920s.  It can be traced to Oliver Wendell Holmes in Bullen v. Wisconsin.

It was slow to be accepted in the United Kingdom. By the 1950s, knowledgeable and careful writers in the UK had come to distinguish the term "tax evasion" from "avoidance". However, in the UK at least, "evasion" was regularly used (by modern standards, misused) in the sense of avoidance, in law reports and elsewhere, at least up to the 1970s. Now that the terminology has received official approval in the UK (Craven v White), this usage should be regarded as erroneous. But even now it is often helpful to use the expressions "legal avoidance" and "illegal evasion", to make the meaning clearer.

Tax Avoidance and Evasion in the United States

In the United States "tax evasion" is evading the assessment or payment of a tax that is already legally owed at the time of the criminal conduct.  Tax evasion is criminal, and has no effect on the amount of tax actually owed, although it may give rise to substantial monetary penalties.

By contrast, the term "tax avoidance" describes lawful conduct, the purpose of which is to avoid the creation of a tax liability in the first place. Whereas an evaded tax remains a tax legally owed, an avoided tax is a tax liability that has never existed.

For example, consider two businesses, each of which have a particular asset (in this case, a piece of real estate) that is worth far more than its purchase price.

• Business One (or an individual) sells the property and underreports its gain. In this instance, tax is legally due. Business One has engaged in tax evasion, which is criminal.
• Business Two (or an individual) consults with a tax advisor and discovers that the business can structure a sale as a "like-kind exchange" (formally known as a 1031 exchange, named after the Code section) for other real estate that the business can use. In this instance, no tax is due of the provisions of section 1031 of the Internal Revenue Code. Business Two has engaged in tax avoidance (or tax mitigation), which is completely within the law.

In the above example, tax may or may not eventually be due when the second property is sold. Whether and how much tax will be due will depend on circumstances and the state of the law at the time.

Definition of Tax Evasion in the United States

The application of the U.S. tax evasion statute may be illustrated in brief as follows. The statute is Internal Revenue Code section 7201:

Any person who willfully attempts in any manner to evade or defeat any tax imposed by this title or the payment thereof shall, in addition to other penalties provided by law, be guilty of a felony and, upon conviction thereof, shall be fined not more than $100,000 ($500,000 in the case of a corporation), or imprisoned not more than 5 years, or both, together with the costs of prosecution.

Under this statute and related case law, the prosecution must prove, beyond a reasonable doubt, each of the following three elements:

1. the "attendant circumstance" of the existence of a tax deficiency – an unpaid tax liability; and

2. the actus reus (i.e., guilty conduct) – an affirmative act (and not merely an omission or failure to act) in any manner constituting evasion or an attempt to evade either:

i.  the assessment of a tax, or

ii. the payment of a tax.

3. the mens rea or "mental" element of willfulness – the specific intent to violate an actually known legal duty.

An affirmative act "in any manner" is sufficient to satisfy the third element of the offense. That is, an act which would otherwise be perfectly legal (such as moving funds from one bank account to another) could be grounds for a tax evasion conviction (possibly an attempt to evade payment), provided the other two elements are also met. Intentionally filing a false tax return (a separate crime in itself)[33] could constitute an attempt to evade the assessment of the tax, as the Internal Revenue Service bases its initial assessment (i.e., the formal recordation of the tax on the books of the U.S. Treasury) on the tax amount shown on the return.

https://en.wikipedia.org/wiki/Tax_noncompliance

Mohs Scale of Mineral Hardness

The Mohs scale of mineral hardness (/moʊz/) is a qualitative ordinal scale, from 1 to 10, characterizing scratch resistance of various minerals through the ability of harder material to scratch softer material.

The scale was created in 1822 by German geologist and mineralogist Friedrich Mohs; it is one of several definitions of hardness in materials science, some of which are more quantitative.

The method of comparing hardness by observing which minerals can scratch others is of great antiquity, having been mentioned by Theophrastus in his treatise On Stones, c.  300 BC, followed by Pliny the Elder in his Naturalis Historia, c.  AD 77. The Mohs scale is extremely useful for identification of minerals in the field, but is not an accurate predictor of how well materials endure in an industrial setting – toughness.

Uses of the Mohs Scale

Despite its lack of precision, the Mohs scale is relevant for field geologists, who use the scale to roughly identify minerals using scratch kits. The Mohs scale hardness of minerals can be commonly found in reference sheets.

Mohs hardness is useful in milling. It allows assessment of which kind of mill will best reduce a given product whose hardness is known.  The scale is used at electronic manufacturers for testing the resilience of flat panel display components (such as cover glass for LCDs or encapsulation for OLEDs).

The Mohs scale has been used to evaluate the hardness of smartphone screens. Most modern smartphone displays use Gorilla Glass that scratches at level 6 with deeper grooves at level 7 on the Mohs scale of hardness.

Minerals Involved in Testing

The Mohs scale of mineral hardness is based on the ability of one natural sample of mineral to scratch another mineral visibly. The samples of matter used by Mohs are all different minerals. Minerals are chemically pure solids found in nature. Rocks are made up of one or more minerals. As the hardest known naturally occurring substance when the scale was designed, diamonds are at the top of the scale. The hardness of a material is measured against the scale by finding the hardest material that the given material can scratch, or the softest material that can scratch the given material. For example, if some material is scratched by apatite but not by fluorite, its hardness on the Mohs scale would fall between 4 and 5.

"Scratching" a material for the purposes of the Mohs scale means creating non-elastic dislocations visible to the naked eye. Frequently, materials that are lower on the Mohs scale can create microscopic, non-elastic dislocations on materials that have a higher Mohs number. While these microscopic dislocations are permanent and sometimes detrimental to the harder material's structural integrity, they are not considered "scratches" for the determination of a Mohs scale number.

The Mohs scale is a purely ordinal scale.  For example, corundum (9) is twice as hard as topaz (8), but diamond (10) is four times as hard as corundum.

Hardness	Substance or mineral
0.2–0.3	caesium, rubidium
0.5–0.6	lithium, sodium, potassium, candle wax
1	talc
1.5	gallium, strontium, indium, tin, barium, thallium, lead, graphite, ice[17]
2	hexagonal boron nitride,[18] calcium, selenium, cadmium, sulfur, tellurium, bismuth, gypsum
2–2.5	halite (rock salt), fingernail[19]
2.5–3	gold, silver, aluminium, zinc, lanthanum, cerium, jet
3	calcite, copper, arsenic, antimony, thorium, dentin
3.5	platinum
4	fluorite, iron, nickel
4–4.5	ordinary steel
5	apatite (tooth enamel), zirconium, palladium, obsidian (volcanic glass)
5.5	beryllium, molybdenum, hafnium, glass, cobalt
6	orthoclase, titanium, manganese, germanium, niobium, uranium
6–7	fused quartz, iron pyrite, silicon, ruthenium, iridium, tantalum, opal, peridot, tanzanite, rhodium, jade
7	osmium, quartz, rhenium, vanadium
7.5–8	emerald, beryl, zircon, tungsten, spinel
8	topaz, cubic zirconia, hardened steel
8.5	chrysoberyl, chromium, silicon nitride, tantalum carbide
9	corundum (includes sapphire and ruby), tungsten carbide, titanium nitride
9–9.5	silicon carbide (carborundum), tantalum carbide, zirconium carbide, alumina, beryllium carbide, titanium carbide, aluminum boride, boron carbide.[a][20][21]
9.5–near 10	boron, boron nitride, rhenium diboride (a-axis),[22] stishovite, titanium diboride, moissanite (crystal form of silicon carbide)
10	diamond, carbonado

 

                       https://en.wikipedia.org/wiki/Mohs_scale_of_mineral_hardness

First Known Black Hole/Neutron Star Mergers

“Elusive missing piece of the family picture of compact object mergers”

From: Northwestern University

June 29, 2021 -- A long time ago, in two galaxies about 900 million light-years away, two black holes each gobbled up their neutron star companions, triggering gravitational waves that finally hit Earth in January 2020. Astrophysicists' observation of the two events -- detected just 10 days apart -- mark the first-ever detection of a black hole merging with a neutron star.

Discovered by an international team of astrophysicists including Northwestern University researchers, two events -- detected just 10 days apart -- mark the first-ever detection of a black hole merging with a neutron star. The findings will enable researchers to draw the first conclusions about the origins of these rare binary systems and how often they merge.

"Gravitational waves have allowed us to detect collisions of pairs of black holes and pairs of neutron stars, but the mixed collision of a black hole with a neutron star has been the elusive missing piece of the family picture of compact object mergers," said Chase Kimball, a Northwestern graduate student who co-authored the study. "Completing this picture is crucial to constraining the host of astrophysical models of compact object formation and binary evolution. Inherent to these models are their predictions of the rates that black holes and neutron stars merge amongst themselves. With these detections, we finally have measurements of the merger rates across all three categories of compact binary mergers."

The research will be published June 29 in the Astrophysical Journal Letters. The team includes researchers from the LIGO Scientific Collaboration (LSC), the Virgo Collaboration and the Kamioka Gravitational Wave Detector (KAGRA) project. An LSC member, Kimball led calculations of the merger rate estimates and how they fit into predictions from the various formation channels of neutron stars and black holes. He also contributed to discussions about the astrophysical implications of the discovery.

Kimball is co-advised by Vicky Kalogera, the principal investigator of Northwestern's LSC group, director of the Center for Interdisciplinary Exploration and Research in Astrophysics (CIERA) and the Daniel I. Linzer Distinguished Professor of Physics and Astronomy in the Weinberg Colleges of Arts and Sciences; and by Christopher Berry, an LSC member and the CIERA Board of Visitors Research Professor at Northwestern as well as a lecturer at the Institute for Gravitational Research at the University of Glasgow. Other Northwestern co-authors include Maya Fishbach, a NASA Einstein Postdoctoral Fellow and LSC member.

Two events in ten days

The team observed the two new gravitational-wave events -- dubbed GW200105 and GW200115 -- on Jan. 5, 2020, and Jan. 15, 2020, during the second half of the LIGO and Virgo detectors third observing run, called O3b. Although multiple observatories carried out several follow-up observations, none observed light from either event, consistent with the measured masses and distances.

"Following the tantalizing discovery, announced in June 2020, of a black-hole merger with a mystery object, which may be the most massive neutron star known, it is exciting also to have the detection of clearly identified mixed mergers, as predicted by our theoretical models for decades now," Kalogera said. "Quantitatively matching the rate constraints and properties for all three population types will be a powerful way to answer the foundational questions of origins."

All three large detectors (both LIGO instruments and the Virgo instrument) detected GW200115, which resulted from the merger of a 6-solar mass black hole with a 1.5-solar mass neutron star, roughly 1 billion light-years from Earth. With observations of the three widely separated detectors on Earth, the direction to the waves' origin can be determined to a part of the sky equivalent to the area covered by 2,900 full moons.

Just 10 days earlier, LIGO detected a strong signal from GW200105, using just one detector while the other was temporarily offline. While Virgo also was observing, the signal was too quiet in its data for Virgo to help detect it. From the gravitational waves, the astronomers inferred that the signal was caused by a 9-solar mass black hole colliding with a 1.9-solar mass compact object, which they ultimately concluded was a neutron star. This merger happened at a distance of about 900 million light-years from Earth.

Because the signal was strong in only one detector, the astronomers could not precisely determine the direction of the waves' origin. Although the signal was too quiet for Virgo to confirm its detection, its data did help narrow down the source's potential location to about 17% of the entire sky, which is equivalent to the area covered by 34,000 full moons.

Where do they come from?

Because the two events are the first confident observations of gravitational waves from black holes merging with neutron stars, the researchers now can estimate how often such events happen in the universe. Although not all events are detectable, the researchers expect roughly one such merger per month happens within a distance of one billion light-years.

While it is unclear where these binary systems form, astronomers identified three likely cosmic origins: stellar binary systems, dense stellar environments including young star clusters, and the centers of galaxies.

The team is currently preparing the detectors for a fourth observation run, to begin in summer 2022.

"We've now seen the first examples of black holes merging with neutron stars, so we know that they're out there," Fishbach said. "But there's still so much we don't know about neutron stars and black holes -- how small or big they can get, how fast they can spin, how they pair off into merger partners. With future gravitational wave data, we will have the statistics to answer these questions, and ultimately learn how the most extreme objects in our universe are made."

Astrophysicists detect first black hole-neutron star mergers: Mix pair is 'elusive missing piece of the family picture of compact object mergers' -- ScienceDaily

Artificial Intelligence Learns to Predict Human Behavior from Analysis of Videos

Computer vision technique leverages higher-level associations between people, animals, and objects

From:  Columbia University School of Engineering and Applied Science

June 28, 2021 -- Predicting what someone is about to do next based on their body language comes naturally to humans but not so for computers. When we meet another person, they might greet us with a hello, handshake, or even a fist bump. We may not know which gesture will be used, but we can read the situation and respond appropriately.

In a new study, Columbia Engineering researchers unveil a computer vision technique for giving machines a more intuitive sense for what will happen next by leveraging higher-level associations between people, animals, and objects.

"Our algorithm is a step toward machines being able to make better predictions about human behavior, and thus better coordinate their actions with ours," said Carl Vondrick, assistant professor of computer science at Columbia, who directed the study, which was presented at the International Conference on Computer Vision and Pattern Recognition on June 24, 2021. "Our results open a number of possibilities for human-robot collaboration, autonomous vehicles, and assistive technology."

It's the most accurate method to date for predicting video action events up to several minutes in the future, the researchers say. After analyzing thousands of hours of movies, sports games, and shows like "The Office," the system learns to predict hundreds of activities, from handshaking to fist bumping. When it can't predict the specific action, it finds the higher-level concept that links them, in this case, the word "greeting."

Past attempts in predictive machine learning, including those by the team, have focused on predicting just one action at a time. The algorithms decide whether to classify the action as a hug, high five, handshake, or even a non-action like "ignore." But when the uncertainty is high, most machine learning models are unable to find commonalities between the possible options.

Columbia Engineering PhD students Didac Suris and Ruoshi Liu decided to look at the longer-range prediction problem from a different angle. "Not everything in the future is predictable," said Suris, co-lead author of the paper. "When a person cannot foresee exactly what will happen, they play it safe and predict at a higher level of abstraction. Our algorithm is the first to learn this capability to reason abstractly about future events."

Suris and Liu had to revisit questions in mathematics that date back to the ancient Greeks. In high school, students learn the familiar and intuitive rules of geometry -- that straight lines go straight, that parallel lines never cross. Most machine learning systems also obey these rules. But other geometries, however, have bizarre, counter-intuitive properties; straight lines bend and triangles bulge. Suris and Liu used these unusual geometries to build AI models that organize high-level concepts and predict human behavior in the future.

"Prediction is the basis of human intelligence," said Aude Oliva, senior research scientist at the Massachusetts Institute of Technology and co-director of the MIT-IBM Watson AI Lab, an expert in AI and human cognition who was not involved in the study. "Machines make mistakes that humans never would because they lack our ability to reason abstractly. This work is a pivotal step towards bridging this technological gap."

The mathematical framework developed by the researchers enables machines to organize events by how predictable they are in the future. For example, we know that swimming and running are both forms of exercising. The new technique learns how to categorize these activities on its own. The system is aware of uncertainty, providing more specific actions when there is certainty, and more generic predictions when there is not.

The technique could move computers closer to being able to size up a situation and make a nuanced decision, instead of a pre-programmed action, the researchers say. It's a critical step in building trust between humans and computers, said Liu, co-lead author of the paper. "Trust comes from the feeling that the robot really understands people," he explained. "If machines can understand and anticipate our behaviors, computers will be able to seamlessly assist people in daily activity."

While the new algorithm makes more accurate predictions on benchmark tasks than previous methods, the next steps are to verify that it works outside the lab, says Vondrick. If the system can work in diverse settings, there are many possibilities to deploy machines and robots that might improve our safety, health, and security, the researchers say. The group plans to continue improving the algorithm's performance with larger datasets and computers, and other forms of geometry.

"Human behavior is often surprising," Vondrick commented. "Our algorithms enable machines to better anticipate what they are going to do next."

              https://www.sciencedaily.com/releases/2021/06/210628113746.htm