Goto

Collaborating Authors

Texas


Using AI to Assist Those Experiencing Homelessness in Austin - AnalyticsWeek

#artificialintelligence

MetroLab Network has partnered with Government Technology to bring its readers a segment called the MetroLab Innovation of the Month Series, which highlights impactful tech, data and innovation projects underway between cities and universities. If you'd like to learn more or contact the project leads, please contact MetroLab at info@metrolabnetwork.org for more information. In this month's installment of the Innovation of the Month series, we explore a collaboration between the University of Texas at Austin and the city of Austin, involving leveraging AI to improve the lives of people experiencing homelessness. MetroLab's Ben Levine spoke with Sherri R. Greenberg from the UT-Austin LBJ School of Public Affairs; Min Kyung Lee, Stephen C. Slota and Kenneth R. Fleischmann from the UT-Austin School of Information; James Snow from the city of Austin Public Works Department; and Jonathan Tomko from the city of Austin Neighborhood Housing and Community Development Department about the background and development of their project. Ben Levine: Can you describe the origin and objective of this project and who has been involved in it?


UTSA Launches Research Center to Expand Reach of Artificial Intelligence

#artificialintelligence

To explore all newsletters, click here. By signing, you agree to the terms of service and privacy policy. Self-driving cars, single-pilot commercial planes, robotic soldiers, and widespread gene editing may still be things of the future, but a new research center in San Antonio is working to bring these and other artificial intelligence innovations to life. The University of Texas at San Antonio officially launched its newest research center, the UTSA Matrix AI Consortium, on Thursday morning via a livestream kickoff event. The consortium will bring together experts studying artificial intelligence to expand the use and deployment of AI. "This initiative is a concerted effort to promote AI innovation, something I'm a big fan about these days," UTSA President Taylor Eighmy said.


Perimeter Medical Imaging Announces Expansion of ATLAS AI Project with Installation of OTISTM for AI development at Leading Cancer Care Center, MD Anderson

#artificialintelligence

DALLAS, TX / ACCESSWIRE / July 27, 2020 / Perimeter Medical Imaging, AI Inc. (TSXV:PINK) today announced the installation of their OTISTM device at the University of Texas MD Anderson Cancer Center (MD Anderson), to further develop ImgAssist AI technology marking an important milestone in this collaboration and Perimeter's ATLAS AI Project. Initiated in mid-July, the ATLAS AI Project allows Perimeter to collaborate with industry-leading cancer care centers that will use OTIS - its proprietary ultra-high resolution imaging platform - to collect images of breast tumors from approximately 400 patients for the purpose of training and testing Perimeter's ImgAssist AI technology. This technology, which is currently under development, is designed to utilize a machine learning model to help surgeons identify, in real-time, if cancer is still present when performing breast-conserving surgery (lumpectomy). This study was made possible, in part, by a $7.4 million grant awarded by the Cancer Prevention and Research Institute of Texas (CPRIT), a leading state body funding cancer research. Jeremy Sobotta, President and CFO stated, "Initiation at MD Anderson is an important milestone in part one of our ATLAS AI Project and marks the next step in our development and clinical validation efforts for our ImgAssist AI software. MD Anderson is one of the largest breast cancer centers in the United States, treating approximately 40,000 patients a year, and is a valued collaborator as we strive to help physicians improve surgical outcomes for breast cancer patients by providing an additional tool for real-time margin visualization and assessment."


Lizard man

Science

For Jonathan Losos, tiny Caribbean islands and their reptile inhabitants are test tubes of evolution. The morning of 17 October 1996 started as usual for Jonathan Losos. The evolutionary biologist donned a broad hat and slathered on sunscreen, then headed by boat to several unnamed islets off Great Exuma Island in the Bahamas. Three years earlier, he and ecologist David Spiller had introduced local lizard species there to learn how they would compete in a once lizardless place. The pair spent the day snaring lizards, noting their exact locations, and taking stock of the insects, spiders, and vegetation. They were worried about reports of an impending hurricane, but the locals seemed confident it would veer off and spare the islands, as usual. Not this time, however. The next day, Losos and Spiller helped their hotel owner board up the windows of their beachfront cottage on Great Exuma as Hurricane Lili bore down on the island. As the wind picked up and the first squalls dumped rain, they scurried to a cinder block building up a hill. That night, the wind blew off parts of the roof and felled palm trees. A 4-meter storm surge flooded the streets, and 2 days later they found their rented motorboat stuck in a tree. The lizards had it even worse. When Losos and Spiller finally made it back out to their most exposed study sites, the islands were stripped nearly bare of brush and all the lizards were gone. But the setback for Losos's project was the start of a new chapter in his research on how the animals adapt to the varied, changeable environments on islands in and around the Caribbean. Since Lili, a half-dozen other hurricanes have inundated islets and swept away animals relocated there by Losos, who is based at Washington University in St. Louis (WashU), and his team. But he and his colleagues have persevered, collecting data on how the animals adapt to predators, storm damage, and other challenges—natural and those contrived by the researchers. A lifelong reptile enthusiast, Losos is driven in part by his passion for a group of lizards called anoles, which thrive in South and Central America and throughout the Caribbean. He also views them as an opportunity. Almost half of the 400 anole species live on islands, and their diverse lifestyles, habitats, and histories have proved to be a vehicle for exploring some of evolution's biggest questions. “Jonathan's islands are like giant test tubes, and he is the ultimate tinkerer,” says Martha Muñoz, an evolutionary biologist at Yale University. Losos's research on anoles has shown that evolution can happen faster than most scientists had assumed, and that—contrary to what some leading thinkers have proposed—it is often predictable. Faced with similar challenges, separate populations often evolve similar solutions. Along the way, Losos has mentored dozens of young scientists, and some are now carrying his work in new directions. “Beyond his many contributions to the field, Jonathan has also changed the course of science simply by being who he is,” Muñoz, a former student, says. “He is proof that success is richer and more rewarding when accompanied by kindness and humility.” ODDLY ENOUGH, THE 1950S TV show Leave it to Beaver started Losos down this path. When 7-year-old Beaver brought home a pet alligator, young Losos asked his parents whether he, too, could get one. His mom was against it, but his father said he would ask a family friend, the deputy director of the St. Louis Zoo, for advice. A successful businessman, the senior Losos also loved animals, taking his family on nature vacations, joining the zoo's board, and even financing the zoo's acquisition of a baby elephant from Thailand, which he named Carolyn in honor of his wife. To everyone's surprise, the director heartily approved, saying that having an alligator as a childhood pet was how he got his start in herpetology. So the junior Losos acquired several baby caimans, which lived in a baby pool in the basement in winter and in a horse trough in the yard the rest of the year. Only a few times did the animals escape and terrorize the neighbors. Losos worked summers at the zoo until partway through college, eventually donating his caimans to a zookeeper. “Jonathan started off as a little kid loving nature, endlessly pestering staff at his local zoo, catching lizards on family vacations, and he's never lost that spark,” says Harry Greene, herpetologist emeritus at Cornell University and Losos's graduate school adviser. As an undergraduate at Harvard University, Losos fell under the tutelage of herpetologist Ernest Williams. Sometimes referred to as the father of anole biology, Williams had recognized that anoles on different Caribbean islands evolved independently. Yet on each island he'd found a similar set of body types or “ecomorphs”—one specialized for living in the brush, another for gripping twigs, and still others for life high in the trees. These parallels suggested that where circumstances were similar, evolution would converge on the same set of traits and form communities with similar sets of species. Williams's lab had already produced several leading evolutionary biologists, and Losos figured the field of anole research was getting too crowded. But no other species both captured his interest and was easy to study. “I went through a dozen failed Ph.D. projects,” he recalls. At a low point, he seriously considered law school, but his dad convinced him that the world needed herpetologists more than lawyers. Losos eventually realized that anoles were perfect for applying new tools in evolutionary biology. Researchers were just beginning to build family trees and trace evolution based on protein variations among species. For his Ph.D., Losos compared proteins in Caribbean anoles and verified that Williams's ecomorphs had indeed evolved independently to form similar communities on different islands ( Science , 27 March 1998, p. [2115][1]). That insight alone—support for an idea called convergent evolution—“was a really important breakthrough,” says Frank Burbrink, a herpetologist at the American Museum of Natural History. Meanwhile, other researchers were calling for more rigor in evolution studies by requiring evidence that supposedly adaptive traits really give an organism an advantage. So Losos began to study different anole ecomorphs, with legs and toepads of varying sizes (see graphic, p. 499). In the lab, he ran them down miniature racetracks and assessed how well they clung to smooth, vertical surfaces. He found that lizards living near the ground, close to predators, had longer legs that made them fast, whereas those living higher in brush and trees had bigger toepads to stick to leaves and smooth bark. By combining these data with his family tree studies, he got a clearer sense of the lizards' evolutionary history. He “was really one of the first people to move the field into doing evolution by integrating ecology and morphology and getting the bigger picture,” Burbrink says. Inspired by experiments in which researchers monitored evolutionary changes in guppies in Trinidad after relocating them to different streams ( Science , 24 August 2012, p. [904][2]), Losos began to wonder whether similar studies could be done in Caribbean anoles. And he realized that Thomas Schoener, one of Williams's protégés, had already laid the groundwork. In the 1980s, Thomas and Amy Schoener (they were once married) introduced local lizards to tiny lizardless islands in the Bahamas to investigate how different vegetation affected the reptiles' ability to thrive. A decade later, Losos teamed up with Thomas Schoener, by then a renowned ecologist at the University of California (UC), Davis, to revisit those sites. Consistent with Williams's and Losos's earlier findings, lizards living in scrubby vegetation had shorter legs and larger toepads than their ancestors, which had lived in tall, broad trees. These adaptations enabled them to cling to tiny twigs as they chased down insects to eat, and the changes had taken just a few generations. “Evolution can happen very quickly when natural selection is very strong,” Losos says. The idea is now well-accepted, but at the time it went against the entrenched belief that evolution was a slow process. “This is one of the few things that [Charles] Darwin got wrong,” Losos says. He decided to make anoles his life's work. HE SOON HAD TO RECKON with hurricanes. Losos and Spiller, now retired from UC Davis, had chosen the islets off Great Exuma to study the effects of competition. On some, they introduced two local species, the green and brown anoles, and on others, just a single species. In the first 3 years, they noticed that on islands with both kinds, the brown lizards were driving the green anoles higher into the bushes, where they were struggling. That's when Lili hit, ruining the experiment before they could see whether the green anoles would go extinct. “It would have been so easy, I'm sure, to pack it all in and give up,” says Luke Harmon, one of Losos's former students and now an evolutionary biologist at the University of Idaho. Instead, Losos and Spiller used the disaster to their advantage. They documented Lili's great, but also patchy, impact. Islands southwest of Great Exuma felt the brunt of the storm surge and were devoid of lizards and vegetation. Life there would have to start over. On islands to the north, the wind and rain snapped twigs and ripped off leaves but a few lizards remained, they reported in the first of several papers about hurricanes ( Science , 31 July 1998, p. [695][3]). The work challenged a widespread assumption that extreme events such as hurricanes do not drive evolution because they are rare and have random, unpredictable impacts on plants and animals. The Losos group discovered instead that storms can be agents of natural selection. For example, in 2017 Losos's postdoc Colin Donihue; functional morphologist Anthony Herrel, now with the French national research agency CNRS at the National Museum of Natural History; and colleagues visited two cays in the Turks and Caicos to measure the body proportions of the anoles living there. Four days after they left, two almost back-to-back hurricanes hit the area with winds of more than 200 kilometers per hour. When the team returned a few weeks later and remeasured the lizards, they found that the survivors tended to have bigger toepads, longer forelimbs, and shorter hindlimbs. Back in the lab, the researchers tested how these traits affect the lizards' ability to hold onto a perch. In a strong wind, anoles hang on with their forelimbs, but they lose their grip with the hind legs. Cranking up an air-blower, the researchers found that those with longer hind legs (and more surface area for the wind to catch) got blown off their perches onto a padded surface more readily. Conversely, animals with shorter hind limbs and bigger toepads hung on. The hurricanes had apparently selected for those traits, the team reported in 2018. The following year, they found that offspring of the survivors also had big toepads, suggesting the adaptation was genetic and not just a reaction to holding on tight. The team has since measured toepad size in 188 lizard species across the Caribbean. The more hurricanes an island has experienced, the bigger the toepads of the lizards living there, they reported on 27 April in the Proceedings of the National Academy of Sciences . Hurricanes seem to have had a long-term evolutionary effect. LOSOS HAD BEEN A PROFESSOR at WashU for 13 years when Harvard came calling in 2005, seeking to recruit him to its evolutionary biology department. A St. Louis native and a hardcore St. Louis Cardinals baseball fan, he hesitated. He even did a yearlong sabbatical at Harvard before finally accepting, in large part because the position included a curatorship at the Harvard Museum of Comparative Zoology. “That was the one thing St. Louis didn't have,” he recalls. There, he continued to build on a reputation for being a kind, enthusiastic mentor. “I have seen him give high school students the same attention and respect that he gives his closest colleagues,” says Melissa Kemp, a former postdoc now at the University of Texas, Austin. “He seems to always be focused on his work, but he also has a whimsical sense of fun at the same time,” says Michele Johnson, a former student and an evolutionary biologist at Trinity University. Losos sports a watch with an anole he photographed as its face and is not above lecturing undergraduates while dressed as a platypus—one of his favorite animals since childhood. Those traits and a firm belief that “there is no one-size-fits-all in terms of how to interact with and mentor students” have helped Losos launch the careers of 59 graduate students and postdocs. They include at least eight Black, Latino, and Native American scholars, in a field that lacks diversity. (Although 3% of U.S. biologists are African American or Black, for example, only 0.3% of evolutionary biologists are.) Ambika Kamath, now a postdoctoral researcher at UC Berkeley, says Losos backed her completely when her studies challenged the long-held idea that male lizards hold territories to corral their mates. She argued instead that females move around and play a role in mate choice. “It would have been much harder for me to do that work without his excitement,” she recalls. Losos worked hard with her to get the paper just right and was eager to be a co-author. “Otherwise it would have just been the work of this young brown woman who could have easily been dismissed as an angry feminist.” Kamath and other students praise Losos for pushing them intellectually without undermining their confidence. Harmon jokes that Losos would never dismiss an idea from his students, no matter how wacky. Instead, he would just pause and say “interesting.” “Eventually I figured out that maybe I should think things through a bit more, if Jonathan thought they were ‘interesting,’” Harmon says. LOSOS AND HIS TEAM keep testing their ideas about ecology and evolution on Caribbean islands. In one recent project, Robert Pringle, now at Princeton University, and Losos tested a key principle in ecology—that introducing a top predator tends to increase biodiversity. The researchers added a predatory ground-dwelling lizard to islands with brown and green anoles. To escape this new threat, the brown anoles began to hang out higher in the foliage, displacing the green anoles that normally lived there and driving them toward extinction. Contrary to conventional wisdom, the predator appeared to be pushing the islands toward lower biodiversity, they reported on 5 June in Nature . Another recent study, led by one of Losos's former postdocs, examined the impacts of an invasive anole species on Dominica. Until 20 years ago, the island was home to a single anole species. Then a lumber shipment introduced a second species that is gradually spreading. To study how the native and invader species interact, behavioral ecologist Claire Dufour, now at the University of Montpellier, used robotic lizards as stand-ins for the invader. The robots did pushups and extended a flap of fake skin under the chin, mimicking the aggressive displays of real lizards. In response, native lizards familiar with the invaders postured more aggressively, suggesting the invaders are forcing the natives to expend more energy defending their territory, the group reported on 27 March in the Journal of Animal Ecology . “Our biggest conclusion is that the species do compete and have negative consequences on each other,” Losos says. ![Figure][4] Evolution's stamp on island-dwelling lizards On islands in and around the Caribbean, 173 species of anole lizards face an array of different environments, predators, and competitors, along with periodic storms. The result is a laboratory of evolution, where scientists have been able to track the speed and course of adaptation. GRAPHIC: V. ALTOUNIAN/ SCIENCE Even as his group continues to churn out papers, Losos is assessing what he has learned so far. In his book Improbable Destinies: Fate, Chance, and the Future of Evolution , published in 2017, he challenged a major contention of one of the field's great thinkers, Stephen Jay Gould, the Harvard paleontologist who argued that chance plays such a big role in determining nature's course that evolution would never take the same path twice. Anoles offer evidence to the contrary, Losos wrote: In similar habitats, they have repeatedly evolved similar body shapes, sizes, and behavior. The book was written for the general public, but it made an impression even on his peers. “I've been studying evolution for 30-plus years, and this book made me rethink some things I thought I knew about biology and evolution,” says Christopher Austin, an evolutionary biologist and herpetology curator at the Louisiana State University Museum of Natural Science. Losos left Harvard in 2018, lured by a new job at WashU and the prospect of returning to his hometown, his cats, and his wife, who has a successful real estate career and did not follow him to Massachusetts. He now heads the Living Earth Collaborative, a biodiversity research initiative that unites experts at the Missouri Botanical Garden, the St. Louis Zoo, and WashU. He is working on a book about evolution in the house cat, another of his favorite species. And he is still dodging hurricanes. Losos and colleagues have been trying to assess the long-term evolutionary impacts of predatory lizards they've introduced to some islands in the Bahamas. “I don't know if we will ever get there,” he says. Every few years a hurricane comes through and blows the evolving lizards away. [1]: http://www.sciencemag.org/content/279/5359/2115 [2]: http://www.sciencemag.org/content/337/6097/904 [3]: http://www.sciencemag.org/content/281/5377/695 [4]: pending:yes


Can we predict solar flares?

Science

Flares from the Sun are the strongest explosions in our Solar System. They can cause severe space weather disturbances, posing a hazard to astronauts and technological systems in space and on the ground. Solar flares have an immediate impact in the form of enhanced radiation and energetic particles in as little as 8 min after the start of the event. Reliable prediction methods for flares are needed to provide longer warning times. However, pinning down the flare onset conditions is necessary for reliable predictions and is still a struggle ([ 1 ][1]). On page 587 of this issue, Kusano et al. ([ 2 ][2]) introduce a method to predict and successfully test for large imminent flares. Since their discovery more than 160 years ago by Carrington and Hodgson ([ 3 ][3], [ 4 ][4]), flares have been associated with sunspots on parts of the Sun with strong magnetic fields called active regions. The vast amount of flare energy is stored in complex (nonpotential) active region magnetic fields. The energy is impulsively released by magnetic reconnection, a fundamental plasma physics process that changes the topology of the magnetic field and converts magnetic energy into kinetic energy, thermal energy, and the acceleration of high-energy particles ([ 5 ][5]). Solar flares have been extensively studied for many decades. Now, regular observations are made at various wavelengths from a large number of ground- and space-based observatories. Regular measurements are made of the magnetic field and its vector components in the photosphere, which is considered the “surface” of the Sun. Despite extensive observations, the specific onset conditions and what triggers a flare are not understood. The lack of magnetic field measurements in the corona, where the field reconfiguration causing the sudden energy release in flares takes place, is a major limitation. Analytical and numerical methods are used instead to reconstruct the coronal magnetic field ([ 6 ][6]), model the instability and evolution of the field using magnetohydrodynamics ([ 5 ][5], [ 7 ][7], [ 8 ][8]), and indirectly infer magnetic reconnection signatures from the spatial and thermal distribution of the flaring plasma ([ 9 ][9]). Magnetohydrodynamic models are an important means for understanding the physics of solar eruptions and their onset conditions, but they cannot be used to predict the time of a flare. Current flare prediction schemes are mostly empirical and based on parameterizations of the surface magnetic field, such as the total magnetic flux of active regions, strong field gradients, and so on ([ 10 ][10], [ 11 ][11]). However, the forecast accuracy of empirical flare prediction schemes is still rather low, as measured by metrics like the skill score. Kusano et al. chose a different, physics-based approach for predicting imminent large flares through a critical condition of a magnetohydrodynamic instability that is triggered by magnetic reconnection. The method of Kusano et al. is based on “double-arc instability” ([ 12 ][12]) that allows a stability assessment of the sigmoidal field that is formed by reconnection between two sheared magnetic flux systems. The instability starts with magnetic reconnection on small spatial scales, called “trigger-reconnection,” between the sheared magnetic loops to create a double-arc loop, which contains the magnetic free energy that can be released during a flare. When the instability grows, the double-arc loops move upward, causing further magnetic reconnection that provides a positive feedback to the instability (see the figure). The double-arc instability is initiated when the ratio of the poloidal flux in the double arc to the flux of the stabilizing overlying magnetic field exceeds a limit. Based on this criterion, Kusano et al. determine the critical length-scale for the trigger-reconnection to destabilize the double arc and estimate the energy that can be released. The predictive model was tested on 200 active regions during solar cycle 24 that contained the largest sunspots but which did not produce large flares, in comparison with all seven active regions that produced large flares above class X2. X-class flares are powerful enough to trigger long-lasting radiation storms. Magnetic field vector measurements of the solar photosphere by the Helioseismic and Magnetic Imager ([ 13 ][13]) onboard NASA's Solar Dynamics Observatory were used for data. The authors found that the location and the time evolution of the identified critical regions provide a precursor to large flares, with lead times of 1 to 24 hours. The two large flares for which this scheme did not work were from a very specific large active region (AR12192) present on the Sun in October 2014. AR12192 was the source of numerous large flares, including six X-class flares, but none of them associated with a large cloud of plasma and embedded magnetic field ejected from the Sun known as a coronal mass ejection ([ 14 ][14]). This is a strong exception to the general statistics, because >90% of all large flares are accompanied by coronal mass ejections ([ 15 ][15]). However, the failure of the Kusano et al. prediction scheme for these events is also relevant because it shows that flare prediction methods have specific difficulties for large flares that are not accompanied by coronal mass ejections. This may be related to flare reconnection occurring relatively high in the corona or to strong overlying fields preventing an eruption ([ 2 ][2], [ 14 ][14]). ![Figure][16] Forecasting solar flares A physics-based model helps better estimate where and when large solar flares (shown above) will erupt. CREDIT: N. CARY/ SCIENCE Predicting solar flares is a very challenging task. The physics is complex and covers a large range of spatial scales, and key observables like the coronal magnetic field are lacking. Finally, the potential that flares are inherently stochastic processes cannot be ruled out. Nonetheless, tackling this issue has occurred by working along different paths. The diverse set of approaches will gain enormously from the upcoming observations of the 4-m Daniel K. Inouye Solar Telescope (DKIST), which had first light in December 2019. DKIST will provide improved resolution of the solar magnetic field fine structure and its dynamics and provide measurements of the coronal magnetic field. These key data are vital for a better understanding and probing of the onset and physics of solar flares. 1. [↵][17]1. L. Green, 2. T. Török, 3. B. VrŠnak, 4. W. Manchester IV, 5. A. Veronig , Space Sci. Rev. 214, 46 (2018). [OpenUrl][18] 2. [↵][19]1. K. Kusano et al ., Science 369, 587 (2020). [OpenUrl][20][CrossRef][21] 3. [↵][22]1. R. A. Carrington , Mon. Not. R. Astron. Soc. 20, 13 (1859). [OpenUrl][23][CrossRef][24] 4. [↵][25]1. R. Hodgson , Mon. Not. R. Astron. Soc. 20, 15 (1859). [OpenUrl][26][CrossRef][27] 5. [↵][28]1. E. R. Priest, 2. T. E. Forbes , Astron. Astrophys. Rev. 10, 313 (2002). [OpenUrl][29] 6. [↵][30]1. C. J. Schrijver et al ., Astrophys. J. 675, 1637 (2008). [OpenUrl][31] 7. [↵][32]1. T. Török, 2. B. Kliem , Astrophys. J. 630, L97 (2005). [OpenUrl][33] 8. [↵][34]1. T. Amari, 2. A. Canou, 3. J.-J. Aly , Nature 514, 465 (2014). [OpenUrl][35] 9. [↵][36]1. Y. Su et al ., Nat. Phys. 9, 489 (2013). [OpenUrl][37] 10. [↵][38]1. C. J. Schrijver , Astrophys. J. 655, L117 (2007). [OpenUrl][39] 11. [↵][40]1. K. D. Leka et al ., Astrophys. J. Suppl. Ser. 243, 36 (2019). [OpenUrl][41] 12. [↵][42]1. N. Ishiguro, 2. K. Kusano , Astrophys. J. 843, 101 (2017). [OpenUrl][43][CrossRef][44] 13. [↵][45]1. P. H. Scherrer et al ., Sol. Phys. 275, 207 (2012). [OpenUrl][46][CrossRef][47] 14. [↵][48]1. J. K. Thalmann, 2. Y. Su, 3. M. Temmer, 4. A. M. Veronig , Astrophys. J. 801, L23 (2015). [OpenUrl][49][CrossRef][50] 15. [↵][51]1. S. Yashiro, 2. S. Akiyama, 3. N. Gopalswamy, 4. R. A. Howard , Astrophys. J. 650, L143 (2005). [OpenUrl][52] Acknowledgments: A.M.V. acknowledges the Austrian Science Fund (FWF): P27292-N20 and I4555-N. [1]: #ref-1 [2]: #ref-2 [3]: #ref-3 [4]: #ref-4 [5]: #ref-5 [6]: #ref-6 [7]: #ref-7 [8]: #ref-8 [9]: #ref-9 [10]: #ref-10 [11]: #ref-11 [12]: #ref-12 [13]: #ref-13 [14]: #ref-14 [15]: #ref-15 [16]: pending:yes [17]: #xref-ref-1-1 "View reference 1 in text" [18]: {openurl}?query=rft.jtitle%253DSpace%2BSci.%2BRev.%26rft.volume%253D214%26rft.spage%253D46%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [19]: #xref-ref-2-1 "View reference 2 in text" [20]: {openurl}?query=rft.jtitle%253DScience%26rft.stitle%253DScience%26rft.aulast%253DKusano%26rft.auinit1%253DK.%26rft.volume%253D369%26rft.issue%253D6503%26rft.spage%253D587%26rft.epage%253D591%26rft.atitle%253DA%2Bphysics-based%2Bmethod%2Bthat%2Bcan%2Bpredict%2Bimminent%2Blarge%2Bsolar%2Bflares%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.aaz2511%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [21]: /lookup/external-ref?access_num=10.1126/science.aaz2511&link_type=DOI [22]: #xref-ref-3-1 "View reference 3 in text" [23]: {openurl}?query=rft.jtitle%253DMon.%2BNot.%2BR.%2BAstron.%2BSoc.%26rft_id%253Dinfo%253Adoi%252F10.1093%252Fmnras%252F20.1.13%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [24]: /lookup/external-ref?access_num=10.1093/mnras/20.1.13&link_type=DOI [25]: #xref-ref-4-1 "View reference 4 in text" [26]: {openurl}?query=rft.jtitle%253DMon.%2BNot.%2BR.%2BAstron.%2BSoc.%26rft_id%253Dinfo%253Adoi%252F10.1093%252Fmnras%252F20.1.15a%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [27]: /lookup/external-ref?access_num=10.1093/mnras/20.1.15a&link_type=DOI [28]: #xref-ref-5-1 "View reference 5 in text" [29]: {openurl}?query=rft.jtitle%253DAstron.%2BAstrophys.%2BRev.%26rft.volume%253D10%26rft.spage%253D313%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [30]: #xref-ref-6-1 "View reference 6 in text" [31]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D675%26rft.spage%253D1637%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [32]: #xref-ref-7-1 "View reference 7 in text" [33]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D630%26rft.spage%253DL97%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [34]: #xref-ref-8-1 "View reference 8 in text" [35]: {openurl}?query=rft.jtitle%253DNature%26rft.volume%253D514%26rft.spage%253D465%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [36]: #xref-ref-9-1 "View reference 9 in text" [37]: {openurl}?query=rft.jtitle%253DNat.%2BPhys.%26rft.volume%253D9%26rft.spage%253D489%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [38]: #xref-ref-10-1 "View reference 10 in text" [39]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D655%26rft.spage%253DL117%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [40]: #xref-ref-11-1 "View reference 11 in text" [41]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%2BSuppl.%2BSer.%26rft.volume%253D243%26rft.spage%253D36%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [42]: #xref-ref-12-1 "View reference 12 in text" [43]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D843%26rft.spage%253D101%26rft_id%253Dinfo%253Adoi%252F10.3847%252F1538-4357%252Faa799b%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [44]: /lookup/external-ref?access_num=10.3847/1538-4357/aa799b&link_type=DOI [45]: #xref-ref-13-1 "View reference 13 in text" [46]: {openurl}?query=rft.jtitle%253DSol.%2BPhys.%26rft.volume%253D275%26rft.spage%253D207%26rft_id%253Dinfo%253Adoi%252F10.1007%252Fs11207-011-9834-2%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [47]: /lookup/external-ref?access_num=10.1007/s11207-011-9834-2&link_type=DOI [48]: #xref-ref-14-1 "View reference 14 in text" [49]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D801%26rft.spage%253DL23%26rft_id%253Dinfo%253Adoi%252F10.1088%252F2041-8205%252F801%252F2%252FL23%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [50]: /lookup/external-ref?access_num=10.1088/2041-8205/801/2/L23&link_type=DOI [51]: #xref-ref-15-1 "View reference 15 in text" [52]: {openurl}?query=rft.jtitle%253DAstrophys.%2BJ.%26rft.volume%253D650%26rft.spage%253DL143%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx


What's This? A Bipartisan Plan for AI and National Security

WIRED

US representatives Will Hurd and Robin Kelly are from opposite sides of the ever-widening aisle, but they share a concern that the US may lose its grip on artificial intelligence, threatening the American economy and the balance of world power. Thursday, Hurd (R-Texas) and Kelly (D-Illinois) offered suggestions to prevent the US from falling behind China, especially, on applications of AI to defense and national security. They want to cut off China's access to AI-specific silicon chips and push Congress and federal agencies to devote more resources to advancing and safely deploying AI technology. Although Capitol Hill is increasingly divided, the bipartisan duo claim to see an emerging consensus that China poses a serious threat and that supporting US tech development is a vital remedy. "American leadership and advanced technology has been critical to our success since World War II, and we are in a race with the government of China," Hurd says. "It's time for Congress to play its role."


Edge AI training on microcontrollers boosts IoT predictive maintenance

#artificialintelligence

One Tech (Dallas, TX) has ported its edge AI unsupervised training technology to microcontrollers to boost analysis in the Internet of Things. MicroAI Atom is designed to be embedded on microcontroller units (MCUs) and can now train and run AI models directly at the endpoint. This enables silicon manufacturers, original equipment manufacturers (OEMs), smart device manufacturers and smart device owners to reduce the costs of bringing intelligence to the edge and endpoint by at least 80 percent. Enabling unsupervised training on a microcontroller at the nertwork edge allows functions such as predictive maintenance that were previously only available on microprocessors. This differs from the AI accelerators on microcontrollers that run an inference engine that has already been trained. "This is a groundbreaking phase for the industry.


Facebook develops AI algorithm that learns to play poker on the fly

#artificialintelligence

Facebook researchers have developed a general AI framework called Recursive Belief-based Learning (ReBeL) that they say achieves better-than-human performance in heads-up, no-limit Texas hold'em poker while using less domain knowledge than any prior poker AI. They assert that ReBeL is a step toward developing universal techniques for multi-agent interactions -- in other words, general algorithms that can be deployed in large-scale, multi-agent settings. Potential applications run the gamut from auctions, negotiations, and cybersecurity to self-driving cars and trucks. Combining reinforcement learning with search at AI model training and test time has led to a number of advances. Reinforcement learning is where agents learn to achieve goals by maximizing rewards, while search is the process of navigating from a start to a goal state.


Coronavirus masks make it harder for facial recognition algorithms to ID people, study finds

FOX News

Protesters in Austin, Texas; Tempe, Ariz.; and Portland, Ore., took the streets in their respective cities to march in step with the Black Lies Matter movement. In some instances, police and federal agents clashed with protesters, but in one demonstration, protesters paid their respects to one of their own in Texas. Coronavirus face masks can confuse facial recognition technology, government researchers announced Monday after a preliminary study on the issue. Facial recognition algorithms developed before the outbreak struggle to identify people wearing masks or face coverings, according to a new study from the U.S. Commerce Department's National Institute of Standards and Technology (NIST). The NIST looked at 89 commercial algorithms and found that even the best had error rates of around 5 percent when trying to match a masked individual's face with their unmasked appearance – up from a normal rate of 0.3 percent when trying to match two photos of the same person.


Texas stadiums helping fight coronavirus with disinfectant-spraying drones

FOX News

The Cotton Bowl is the first stadium in Texas to take a chance on the technology, which has the capability of disinfecting a 92,000 person stadium within 4 hours. DALLAS -- Stadiums are looking for ways to bring fans back to the stands in time for fall sports despite the coronavirus outrbreak, leading some Texas facilities to turn to drones for help. Cotton Bowl senior marketing director Julian Bowman describes the feeling of seeing the iconic Cotton Bowl Stadium in Dallas empty for the last few months, saying, "It is a weird feeling." "The Cotton Bowl opened up in 1930, so this was our 90th year and it was set to be our best year ever and unfortunately with COVID we are not able to do that," Bowman said. "It has really affected how we have been able to connect with our sports community and our entertainment community." The last event the Cotton Bowl was able to host was in January of 2020, before COVID-19 shut them down.