Baidu releases quantum machine learning toolkit on GitHub

Chinese tech giant’s Paddle Quantum development toolkit now is available on GitHub, enabling developers to build and train quantum neural network models, and includes quantum computing applications.

Eileen Yu

By Eileen Yu for By The Way | May 27, 2020 — 16:16 GMT (09:16 PDT) | Topic: Artificial Intelligence

Baidu has released the toolkit for its quantum machine learning platform, Paddle Quantum, which it says will enable developers to build and train quantum neural network models. Built on the Chinese tech giant’s deep learning platform PaddlePaddle, the toolkit also includes quantum computing applications.

Paddle Quantum, currently available on GitHub, comprises a set of quantum machine learning toolkits, including a quantum chemistry library and optimisation tools, as well as three quantum applications: quantum machine learning, quantum chemical simulation, and quantum combinatorial optimisation. 

Several underlying functions of PaddlePaddle, including matrix multiplications, also enable Paddle Quantum to support quantum circuit models and general quantum computing research, Baidu said in a statement on Wednesday.

China can help drive global progress in quantum computing

China can help drive global progress in quantum computing

Asian country has began investing in quantum technology and is at a similar starting point with other economic powers in this field, says Shanghai-born Turing Award winner Andrew Yao.

Touting the platform’s added flexibility, the Chinese tech vendor said Paddle Quantum could run a new implementation of the Quantum Approximate Optimisation Algorithm (QAOA) at half the number of layers in its quantum neural network. 

Director of Baidu’s Institute for Quantum Computing, Duan Runyao, said: “Researchers in the quantum field can use the Paddle Quantum to develop quantum artificial intelligence (AI) and deep-learning enthusiasts have a shortcut to learning quantum computing.”

Baidu also introduced seven new tools that offering 27 enhanced features for PaddlePaddle. These included Paddle.js, a deep learning JapaScript library that would enable developers to use AI within the browser or smart mini programmes in apps such as Wechat. The updates also included Parakeet, a text-to-speech toolkit with various models such as WaveFlow and ClariNet, as well as Paddle X, an integration development tool for data processing. 

According to Baidu, PaddlePaddle has been adopted by more than 1.9 million developers, with more than 84,000 businesses using the deep learning platform to create more than 230,000 models. The company said it also worked with several global hardware manufacturers including Intel, Huawei, MediaTek, and Inspur, on the PaddlePaddle ecosystem. 

For the calendar and fiscal year 2017, it looks like the majority of IT budgets are larger than last year. In an online survey conducted by Tech Pro Research in July, 61% of respondents said their company’s budget will increase. Most in that category…Research provided by TechRepublic Premium

Baidu CTO Wang Haifeng said: “Now is an unprecedented opportunity for the development of PaddlePaddle given the rise of industrial intelligence and the acceleration of AI-powered infrastructure. We will continue to embrace the open-source spirit, drive technological innovation, and partner with developers to advance deep learning and AI technologies and speed up the process of industrial intelligence.”

Alibaba last November also published the core codes of its machine learning platform Alink on GitHub, uploading a rang of algorithm libraries that it said supported batch and stream processing. These were essential to support machine learning tasks such as online product recommendations and smart customer services.

This is what Trinity Bellwoods looks like on its first weekend with social distancing circles

Stay in the loop

Sign up for our free email newsletter. Unsubscribe anytime or contact us for details.ShareTweetSave this Article

It’s the first weekend in Trinity Bellwoods Park where visitors can test out the new social distancing circles at one of Toronto’s most popular green spaces, and for all intents and purposes, it’s been a hit. 

The white circles, which were painted by City of Toronto staff earlier this week as a pilot project, did their job in keeping groups of people isolated from one another while enjoying the outdoors. 

trinity bellwoodsSo far, the park has been significantly less busy than past weekends, where sunny skies and high temperatures lured people out in dangerous crowds. 

trinity bellwoods parkSaturday proved to be a lot chillier than other days we’ve seen, but people seemed eager to test drive the social distancing circles which were popularized in New York and San Francisco before being implemented in Toronto.

trinity bellwoodsAround eight feet in diameter, and spaced about 10 feet apart, each circle is large enough for two adults from the same household lying down, or three adults from the same household if they’re all sitting cross-legged.

trinity bellwoods parkBike cops and bylaw officers were in full effect on Saturday, and seemed much more effective than last week in directing people how to properly physically distance. 

trinity bellwoods park

Lead photo by 

Fareen Karim

Extreme night owls: ‘I can’t tell anyone what time I go to bed’

What happens when your natural sleeping pattern is at odds with the rest of the world?

illustration of woman standing over cutouts of sleeping town

Up all hours: ‘natural night owlsare fundamentally different to insomniacs or people who stay up until the early hours because of family or work circumstances.’ Illustration: Eiko Ojala/The ObserverRachel Hall@rachela_hallPublished onSun 31 May 2020 13.00 BST

or as long as she can remember, Jenny Carter has gone to bed late and not woken up until late the following morning, sometimes even the early afternoon. Growing up, she didn’t have a bedtime, and at university she preferred to write her essays between 6pm and 10pm. She loves evenings. They’re when she feels the most creative and can concentrate the best. But that’s not when her employer or society expects her to be productive.Advertisement

“Going to bed at a ‘normal’ time feels so unnatural to me,” she says. “But society just doesn’t cater for people whose sleep cycle doesn’t fit the generic 9 to 5.” She has got into trouble at work for her timekeeping, which has led to disciplinary action. “I’ve had to write off so many events, meetings and opportunities, because they were in the morning and I just knew I wouldn’t be awake.”

Carter, 27, an NHS co-ordinator, is an “extreme night owl”, one of an estimated 8.2% of the population whose natural inclination is to fall asleep well after midnight. Left to her own devices, she’d prefer to go to bed around 3am and wake up about noon.

She has struggled to organise her life in a way that suits her natural sleeping pattern. She negotiated a slightly later start time at work – 10am – but wishes she could begin at noon and finish at 8pm. Instead, she deprives herself of sleep during the week and catches up at weekends, when she often sleeps until 3pm.Advertisement

But this isn’t what frustrates her most about being a night owl. “I think one of the worst things is people equating night owls and late risers with laziness,” she says. “I am just as productive, enthusiastic and organised as others, but at a different time. Feeling completely out of sync with the rest of society is the hardest thing, like you must be the one that’s wrong.”

There’s a growing body of evidence that suggests it’s society, not night owls like Carter, that is wrong. The field of chronobiology seeks to understand how individuals are driven by an internal clock – their “chronotype” – one that is set by genetics, not willpower. The term night owl is shorthand for the chronotype that drives people to go to bed later and rise later. This contrasts with morning larks, who naturally want to go to bed early and wake up early. Most people fall somewhere between the two, with an average sleep cycle running from around 11.30pm until 7.30am. People tend to change over their lifetime. They are larks in childhood, night owls as teens, and more lark-like again as they get older.

‘Say goodbye to your inner farmer. You don’t have to get up with the cows.’

‘Say goodbye to your inner farmer. You don’t have to get up with the cows.’ Illustration: Eiko Ojala/The ObserverAdvertisement

These preferences have a huge influence on health and wellbeing. Experiments show that teens with later school start times achieve better grades, while adults tend to be healthier and more productive when they are allowed to sleep when they want and to work flexibly.

So why do night owls exist? There is no single universally accepted theory, but evolutionary biologists think that communities with more variation in chronotypes may have been more likely to survive. If not everyone needs to sleep at the same time, then some members of the tribe can stand guard and protect those who are resting.

A recent study of a modern-day hunter-gatherer tribe found that during a three-week period, there were only 18 minutes during which all of the 33 tribe members were asleep simultaneously.

Another theory is that variation is simply how genetics works. Colin Espie, professor of sleep medicine at the University of Oxford, says this mirrors differences in hair, eye and skin colour, or height. “It’s a bit like any other biological characteristic. There’s a normal distribution, so there are people on both extremes – and the majority of people are neither.”

Natural night owlsare fundamentally different to insomniacs or people who stay up until the early hours because of family or work circumstances. Being a night owl isn’t a problem – unless you’re trying to fit into a schedule that doesn’t suit your natural cycle.Advertisement

But this isn’t always well understood. Jessica Batchelor is a medical writer who feels most productive at 11pm in the evening. “I can’t tell anyone when I went to sleep, woke up, showered, ate a meal, or took a nap without being judged,” she says. “I struggle with feelings of guilt and shame.

“We’re brainwashed to believe that early birds are happier, more successful, more disciplined and all-round better human beings than night owls. The hours when I feel most alive are considered ‘ungodly’ and likened to a vampire’s schedule. Owls like myself internalise this message, and we believe we must be lazy, depressed and irresponsible.”

Espie has treated night owls who want to adapt their sleeping patterns. He does this by asking night owls to gradually shift their sleeping pattern earlier, usually by 15 minutes or half an hour per week, through doses of bright light in the morning. This causes the brain to shut down the production of melatonin, the hormone that causes sleepiness. In the evening, night owls must make their bedrooms as dark as possible.

“It will take quite some weeks to shift the body clock, but that way you have the best chance of shifting it for a decent amount of time,” Espie says. “You’re unlikely to convert an extreme night owl to a morning person, but you can help them get to a middle ground.”Advertisement

Espie’s approach is similar to that advocated by academics at three universities whose research received widespread media coverage last year. They showed that night owls can “retrain” their body clocks through lifestyle adjustments, including exercise and meals at set times, combined with light exposure.

This doesn’t necessarily mean that night owls can easily adjust. Many think that the ideal solution for night owls is to find work that better suits their natural rhythm. For some, this may be shift work, while others opt for flexible employment, such as setting up their own business or freelancing.

This is Mary McCleod’s experience. She left her role as a fashion buyer, working 9am to 5pm, to start up her own business selling natural soaps, working 11am until 3pm, then again between 8pm and 1am. “When I was going into the office for 9am I would find my mornings would be fairly unproductive and I would tend to stay late to get my ‘good’ work done, so overall I was missing out on other activities in my day,” she says. “I love working to a schedule that suits me better.”

Hannah Edwards, who runs her cake business after her children have gone to bed, agrees. “I am absent-minded and easily distracted during the day,” she says. “Staying up late to get work finished is never a chore or a challenge – when everyone else is tired I’m just getting going. It means my productivity, creativity and output levels are then incredibly high compared to others who have nothing left in the tank.”Advertisement

Flexible work schedules are currently not the norm, but sleep experts believe they should be. For 15 years, Camilla Tring has run B Society, which advises companies around the world on how to implement “chronoleadership” – the idea that they should adapt their work patterns to suit the sleeping schedules of their employees, rather than the other way around.

While morning larks equally benefit from being allowed to arrive at work early and leave early, Tring feels that the battle is hardest for night owls, who experience more stigma. Many night owls say they receive sarcastic comments from colleagues about being lazy when they arrive at work later, even when they stay late to compensate.

This mentality is rooted in our agrarian past, when farm work had to begin at dawn, she says, since people who slept in were unable to provide for their families. These ingrained belief systems are evidenced through aphorisms that span cultures, such as “the early bird catches the worm”. Tring thinks that they no longer apply to the modern world: “We should say goodbye to our inner farmer – we don’t have to get up with the cows any more.”

Equally, Tring views the idea that everyone should be in the office at the same time as a hangover from the industrial revolution, when most people worked in factories. “It’s this mentality of ‘I see you therefore you’re working,’” she says. “But that’s not the case when you’re a knowledge worker. It’s more about the quality of your work than how much you’re working.”Advertisement

For night owls,the expectation that they should adapt their behaviours can also be frustrating outside work. Lisa Akker, who is 60, has been a night owl her whole life and thinks the tendency runs in her family. “It’s caused problems in my marriage. My husband doesn’t understand why I can’t change my sleep times so I can be more of a lark. My cousin is always making fun of my sleeping late proclivity; she’s the quintessential illustration of the early bird getting the worm, and prizes that.”

Some night owls say that they deliberately schedule emails to send in the morning or avoid texting friends at night for fear of judgment. Akker said she recently messaged a friend at 11.19pm to ask about her coronavirus recovery. “I texted her, saying: ‘I know you’ll answer me tomorrow, but I just wanted to know how you are.’ She responded at 8am: ‘Wow, our sleep schedules are so different!’ And the purpose of saying that is? Jeez, it was only 11.19pm.”

The prevalence of this attitude is especially surprising when you consider that in society we are all becoming more nocturnal. Till Roenneberg, the circadian biologist who developed the Munich ChronoType Questionnaire, used to determine whether we are night owls or morning larks, says that because we are exposing our bodies to less natural light than before, we are shifting our body clocks later.Advertisement

“The biological clock evolved to get a lot of light during the day and get none after night, because we didn’t have electrical light,” Roenneberg says. “In the past, the distribution of larks and owls was much narrower. If you couldn’t fall asleep in those days until 2am and would routinely sleep until 10am you were probably an outlier, or you were sick.”

This is why, according to Roenneberg, it’s now the norm for people to use alarm clocks. He coined the term “social jetlag” to refer to the mismatch between most people’s body clocks and the schedules imposed on them by society. “There is practically no health factor that has been looked at that does not become worse with increasing social jetlag.”

Roenneberg’s vision of an ideal society would see nobody use an alarm clock: “[People] would fall asleep when they’re tired and wake up when they have slept to the biological end.”

So if all thescientific evidencesupports the idea of flexible schedules aligned with our individual sleeping patterns, why aren’t we there yet? Employers have long been hesitant to allow their employees to work flexibly and remotely, though this attitude has been undermined by the sudden shift to online working that the coronavirus pandemic has required.Advertisement

Paul Kelley, an academic who has written a book on sleep, Body Clocks: The Biology of Time for Sleep, Education and Work, thinks the problem is employers’ inherent conservatism. “Let people have a choice and see what happens,” he says. “It doesn’t cost any money and it improves the working of society. The best time you can all get together is early afternoon.”

Shifting to flexible working hours as a result of the coronavirus pandemic has been a revelation for publicist Mayah Riaz. She has always known she preferred to work in the evening, but previously stopped herself from doing so because she worried about being perceived as immature or a workaholic.

“It’s proven to me I do better work in those late hours,” she says of the recent change. “There really is something about the magic of the night and stillness. I need to stop feeling like I ‘should’ wake up at 7am or 8am.”

Part of the reason for this prejudice is that sleep science is the missing piece of the public health puzzle. Our culture mistakenly associates sleeping little and rising early with virtue. It is often extolled as a habit of successful people: for instance in the fascination with Margaret Thatcher’s four-hour rest, or articles about “sleepless-elite” CEOs who start their days with a 4am jog. Yet this belies a glaring inconsistency: around eight hours of good-quality sleep is essential for better health for almost everybody.Advertisement

But Professor Espie thinks the tide is turning. When he first qualified as a doctor 40 years ago, exercise, smoking and diet weren’t taken seriously by the medical profession. “No self-respecting doctor would run clinics encouraging people to lose weight or get fit,” he says. “For many years, we’ve been advocating for sleep and it’s been falling on deaf ears. But nowadays there’s a lot being written about sleep in the media and it’s coming into the public consciousness. That’s hugely important.”

As well as a need for greater awareness of the importance of sleep, there are too many myths circulating around it. Among these is the idea that there’s a simple solution to night owls’ struggles: they should cut down on caffeine, practise better sleep hygiene, be more disciplined – and even rely on medication.

For people like Jenny Carter, this lack of understanding is interfering with her ability to live life to its fullest. She has approached GPs in search of treatment, only to be told that there’s nothing wrong with her and she should just go to bed earlier.

“I don’t necessarily feel like my sleeping pattern is a problem. It’s not that I have trouble sleeping when I am asleep. It’s not that I wake up or struggle to sleep, it’s just I sleep a lot later and wake up a lot later. The idea of taking sleeping pills is weird for me when my actual sleep is fine,” she says. “The problem is more that it just doesn’t fit in with the rest of the world.”

Bats, Pangolins and Humans: COVID-19 Virus Likely Emerged From Recombination of Viral Genes Across Different Species

TOPICS:American Association For The Advancement Of ScienceCOVID-19GeneticsVirology


Genetic Analysis Concept

Emergence of SARS-CoV-2 (the virus that causes COVID-19) through recombination and strong purifying selection.

A combination of genetic shuffling and evolutionary selection of near-identical genetic sequences among specific bat and pangolin coronaviruses may have led to the evolution of SARS-CoV-2 and its introduction into humans, a new study suggests.

The results also showed that the virus’ entire receptor binding motif (RBM), a component that plays a key role in viral entry into host cells, was introduced through recombination with pangolin coronaviruses.

The study joins ongoing efforts to identify the source of the virus that causes COVID-19, which is critical for informing efforts to establish proper animal models, discover new drugs and vaccines, and ultimately prevent the rise of future zoonotic diseases.

While the precise origin of SARS-CoV-2 remains a mystery, this study makes clear “that reducing or eliminating direct human contact with wild animals is critical to preventing new coronavirus zoonoses in the future,” the authors say. Proximity of different species in a wet market setting, for example, may increase the potential for cross-species spillover infections, by enabling recombination between more distant coronaviruses and the emergence of mutations, the authors say.

By analyzing 43 complete genome sequences from three strains of SARS-CoV-2-like coronaviruses from bats and pangolins, Xiaojun Li and colleagues delineated which strains were most and least similar to the novel coronavirus, with a special focus on genes related to the virus’ spike protein complex, a critical component that facilitates viral entry into host cells. They found evidence of strong evolutionary selection around the RBM — part of the spike’s amino acid sequence that directly contacts host cell receptors — among the bat, pangolin, and human coronaviruses they studied. Amino acid sequences from these viruses and SARS-CoV-2 were identical or nearly identical in the regions adjacent to the RBM, suggesting that common evolutionary mechanisms shaped these distinct viral strains.

The scientists also demonstrated that SARS-CoV-2’s entire RBM was introduced through recombination with coronaviruses from pangolins. Together, evolutionary selection and frequent recombination among coronaviruses from bats, pangolins, and humans may have allowed the closely related viruses to readily jump between species, the authors postulate, leading to the introduction of SARS-CoV-2 in humans.

Reference: “Emergence of SARS-CoV-2 through recombination and strong purifying selection” by Xiaojun Li, Elena E. Giorgi, Manukumar Honnayakanahalli Marichannegowda, Brian Foley, Chuan Xiao, Xiang-Peng Kong, Yue Chen, S. Gnanakaran, Bette Korber and Feng Gao, 29 May 2020, Science Advances.
DOI: 10.1126/sciadv.abb9153

Nanoparticles can make home refrigeration more accessible for low-income households



Power consumption of a home refrigerator can be cut by 29% while improving cooling capacity. Researchers replaced widely-used, but environmentally unfriendly, R134a refrigerant with the more energy-efficient R600a. They dosed R600a with multi-walled carbon nanotube (MWCNT) nanoparticles. Drop-in refrigerant replacement in the field by trained technicians is possible, says an engineer from the University of Johannesburg.

This test of nanoparticle-dosed refrigerants is a first of its kind and recently published in Energy Reports, an open-access journal. The results can help make home refrigeration more accessible for low-income families.

R134a is one of the most widely-used refrigerants in domestic and industrial refrigerators. It is safe for many applications because it is not flammable. However, it has high global warming potential, contributing to climate change. It also causes fridges, freezers and air-conditioning equipment to consume a lot of electrical energy. The energy consumption contributes even more to climate change.

Meanwhile, a more energy-efficient refrigerant can result in much lower electricity bills. For vulnerable households, energy security can be improved as a result. Improved energy economy and demand-side management can also benefit planners at power utilities, as cooling accounts for about 40% of energy demand.Nanoparticles enhance power reduction

Nano eco-friendly refrigerants have been made with water and ethylene glycol. Previous studies showed reduced energy use in nano-refrigeration, where refrigerants were dosed with multi-walled carbon nanotube (MWCNT) nanoparticles. The nanoparticles also resulted in reduced friction and wear on appliance vapour compressors.

But previous research did not test the effects of MWCNT’s on hydro-carbon refrigerants such as R600a.

In a recent study, researchers at the University of Johannesburg tested the drop-in replacement of environmentally-unfriendly refrigerant R134a, in a home refrigerator manufactured to work with 100g R134a.

They replaced R134a with the more energy-efficient refrigerant R600a, dosed with MWCNT nanoparticles.Reduces electricity use by more than a quarter

The researchers removed the R134a refrigerant and its compressor oil from a household fridge. They used a new refrigerant, R600a, and dosed it with multi-walled carbon nanotubes (MWCNTs). Mineral oil was used as a lubricant. The new mix was fed into the fridge and performance tests were conducted.

They found that the R600a-MWCNT refrigerant resulted in much better performance and cooling capacity for the fridge.

“The fridge cooled faster and had a much lower evaporation temperature of -11 degrees Celsius after 150 minutes. This was lower than the -8 degrees Celsius for R134a. It also exceeded the ISO 8187 standard, which requires -3 degrees Celsius at 180 minutes,” says Dr Daniel Madyira.

Dr Madyira is from the Department of Mechanical Engineering Science at the University of Johannesburg.

“Electricity usage decreased by 29% compared to using R134a. This is a significant energy efficiency gain for refrigerator users, especially for low income earners,” he adds.

To gain these advantages, the choice of MWCNT nanoparticles is critical, he says.

“The MWCNT’s need to have nanometer-scale particle size, which is extremely small. The particles also need to reduce friction and wear, prevent corrosion and clogging, and exhibit very good thermal conductivity,” says Dr Madyira.Managing flammability

The new refrigerant mix introduces a potential risk though. Unlike R134a, R600a is flammable. On the other hand, it is more energy efficient, and it has a low Global Warming potential. Some refrigerator manufacturers have already adopted production with R600a and these appliances are available in the market.

“To do a safe drop-in replacement, no more than 150g of R600a should be used in a domestic fridge,” says Dr Madyira. “Before the replacement, the fridge used 100g of R134a gas. We replaced that with 50g to 70g of R600a, to stay within safety parameters.”

An untrained person should not attempt this drop-in replacement, says Dr Madyira. Rather, a trained refrigeration technician or technologist should do it.Replacement procedure

“Mineral oil is used as the compressor oil. This should be mixed with the recommended concentration. A magnetic stirrer and ultrasonicator are needed to agitate and homogenize the ingredients in the mixture. The mixture can then be introduced into the compressor. After that, R600a can be charged into the refrigerator compressor, while taking care to not use more than 150g of the gas,” says Dr Madyira.A woman’s fridge is her castle

A far more energy-efficient refrigerant, such as the R600a-MWCNT mix, can save consumers a lot of money. Vulnerable households in hot climates in developing countries can benefit even more.

Low income earners in many countries are dependent on home fridges and freezers to safely store bulk food supplies. This greatly reduces the risk of wasting food due to spoilage, or food poisoning due to improperly stored food. These appliances are no longer a luxury but a necessity, says Dr Madyira.

Without fridges, people may be forced to buy food daily in small quantities and at much higher prices. Because daily buying may not be required anymore, travel time and costs for buying food can be much lower as well.

Refrigeration also makes it possible to safely store more diverse food supplies, such as fresh fruit and vegetables. Medicines that require cooling can be stored at home. This can make more balanced diets and nutrition, and better physical health, more accessible for a low-income household.Grid power still rules for low-income refrigeration

From a sustainability point of view, it can look preferable to run most home fridges and freezers from solar power.

However solar panels, backup batteries, and direct current (DC) fridges are still too expensive for most low-income families in areas served by power utilities.

Energy-efficient, alternating current (AC) fridges running on grid power may be more affordable for most. Further cutting power consumption with R600a-MWCNT refrigerant can bring down costs even more.Refrigeration for all vs demand-side management

As more low-income households and small businesses switch on grid-powered fridges, freezers and air-conditioning, power demand needs be managed better.

In South Africa where the study was conducted, the state-operated power utility faces huge challenges in meeting demand consistently. Long-lasting rolling blackouts, known as load-shedding, have been implemented as a demand-side power management measure.

Shaving off more than a quarter of the power consumption of fridges, freezers and air-conditioning units can free up national power supply for improved energy security.

Controlling Two-Dimensional Superconductors With Terahertz Light

TOPICS:2D MaterialsInstitute For Basic ScienceSuperconductivity


Controlling Superconductors With Light Concept

New theoretical study shows how to use Terahertz light to peep in the secrets of two-dimensional superconductors

A researcher at the Center for Theoretical Physics of Complex Systems, within the Institute for Basic Science (IBS, South Korea), Professor Ivan Savenko, has reported a conceptually new method to study the properties of superconductors using optical tools. The theory was published in Physical Review Letters and co-authored by Doctor Vadim Kovalev, physicist at the A.V. Rzhanov Institute of Semiconductor Physics (Russia).

Below some temperature, the resistivity of a material can disappear, and thus, superconducting properties emerge. These are usually extremely low temperatures, between -200°C and -272°C, where commonly unbound electrons suddenly change their behavior and pair up, forming Cooper pairs. This transition manifests itself with supercurrents, which can circulate in the material forever without losses.

Graphic of the system: a two-dimensional superconductor close to its critical temperature. The system is exposed to electromagnetic field (orange arrow) with THz frequency. Stripes of metals as a grating (not shown) are required for the excitation of plasmons, a special type of excitations of unbound electrons in the superconductor. Thus, unbound electrons act as mediators: they interact with each other, with light (as plasmons), and with the Cooper pairs, depicted as dashed red circles. Credit: IBS

However, superconducting properties can appear slightly above the critical temperature. In this so-called fluctuating regime, Cooper pairs start to appear and disappear, drastically altering the electric conductivity and other properties of the superconductor. More than fifty years ago, Aslamazov and Larkin developed a theory which says that the conductivity of fluctuating superconductors is mediated by both unbound electrons and Cooper pairs. However, fluctuating superconductivity is such a challenging research topic that it continues to be investigated. In this new study, the researchers suggest a way to monitor these electron transport phenomena with optical spectroscopy, an experimentally available optical platform.

“While the resistivity-based and magnetic methods to monitor superconductors are well established, it is very hard to “marry” light and superconductivity,” explains Savenko. “This is a hot research field where we can expect new discoveries in fundamental science and innovative applications.”

Superconductivity and light are two seemingly unrelated phenomena. Usually, superconductors are not very sensitive to external light: they can only weakly interact with it, and rather serve as mirrors. This study, instead, shows that light at terahertz (THz) frequencies, which lie between the radio and infrared domains, could be used to optically explore the properties of superconductors.

The researchers modelled the optical and electrical responses of a 2D fluctuating semiconducting layer exposed to THz waves. Approaching the critical temperature, the emerging Cooper pairs cause significant changes in electric conductivity and light absorption by the system. The unbound electrons act as mediators, interacting with both Cooper pairs and light.

“The design we developed is very simple. Therefore, we believe that our discovery can be applicable to multiple cases,” says Savenko. “We expect that the corresponding experiment will be conducted in the near future. It should show either the modification of the electric current, or the alteration of the reflected or transmitted light spectrum, depending on the density of the Cooper pairs.”

Reference: “Proposal for Plasmon Spectroscopy of Fluctuations in Low-Dimensional Superconductors” by V. M. Kovalev and I. G. Savenko, 18 May 2020, Physical Review Letters.
DOI: 10.1103/PhysRevLett.124.207002

Microsoft Just Built a World-Class Supercomputer Exclusively for OpenAI

By Jason Dorrier -May 31, 2020232

Last year, Microsoft announced a billion-dollar investment in OpenAI, an organization whose mission is to create artificial general intelligence and make it safe for humanity. No Terminator-like dystopias here. No deranged machines making humans into paperclips. Just computers with general intelligence helping us solve our biggest problems.

A year on, we have the first results of that partnership. At this year’s Microsoft Build 2020, a developer conference showcasing Microsoft’s latest and greatest, the company said they’d completed a supercomputer exclusively for OpenAI’s machine learning research. But this is no run-of-the-mill supercomputer. It’s a beast of a machine. The company said it has 285,000 CPU cores, 10,000 GPUs, and 400 gigabits per second of network connectivity for each GPU server.

Stacked against the fastest supercomputers on the planet, Microsoft says it’d rank fifth.

The company didn’t release performance data, and the computer hasn’t been publicly benchmarked and included on the widely-followed Top500 list of supercomputers. But even absent official rankings, it’s likely safe to say its a world-class machine.

“As we’ve learned more and more about what we need and the different limits of all the components that make up a supercomputer, we were really able to say, ‘If we could design our dream system, what would it look like?’” said OpenAI CEO Sam Altman. “And then Microsoft was able to build it.”

What will OpenAI do with this dream-machine? The company is building ever bigger narrow AI algorithms—we’re nowhere near AGI yet—and they need a lot of computing power to do it.

The Pursuit of Very Large AI Models

The size of the most advanced AI models—that is, the neural networks in machine learning algorithms—has been growing fast. At the same time, according to OpenAI, the computing power needed to train these models has been doubling every 3.4 months.

The bigger the model, the bigger the computer you need to train it.

This growth is in part due to the number of parameters used in each model. Simplistically, these are the values “neurons” operating on data in a neural net assume through training. OpenAI’s GPT-2 algorithm, which generated convincing text from prompts, consisted of nearly 1.5 billion parameters. Microsoft’s natural language generating AI model, Turing NLG, was over 10 times bigger, weighing in at 17 billion parameters. Now, OpenAI’s GPT-3, just announced Thursday, is reportedly made up of a staggering 175 billion parameters.

There’s another trend at play too.

Whereas many machine learning algorithms are trained on human-labeled data sets, Microsoft, OpenAI, and others are also pursuing “unsupervised” machine learning. This means that with enough raw, unlabeled data the algorithms teach themselves by identifying patterns in that data.

Some of the latest systems can also perform more than one task in a given domain. An algorithm trained on the raw text of billions of internet pages—from Wikipedia entries to self-published books—can infer relationships between words, concepts, and context. Instead of being able to do only one thing, like generate text, it can transfer its learning to multiple related tasks in the same domain, like also reading documents and answering questions.

The Turing NLG and GPT-3 algorithms fall into this category.

“The exciting thing about these models is the breadth of things they’re going to enable,” said Microsoft Chief Technical Officer Kevin Scott. “This is about being able to do a hundred exciting things in natural language processing at once and a hundred exciting things in computer vision, and when you start to see combinations of these perceptual domains, you’re going to have new applications that are hard to even imagine right now.”

If Only We Had a Bigger Computer…

To be clear, this isn’t AGI, and there’s no certain path to AGI yet. But algorithms beginning to modestly generalize within domains is progress.

A looming question is whether the approach will continue progressing as long as researchers can throw more computing power at it, or if today’s machine learning needs to be augmented with other techniques. Also, if the most advanced AI research requires such prodigious resources, then increasingly, only the most well-heeled, well-connected private organizations will be able to play.

Some good news is that even as AI model size is growing, the efficiency of those models is improving too. Each new breakthrough requires a big jump in computing power, but later models are tweaked and tuned, such that successor algorithms can do as well or better with less computing power.

Microsoft also announced an update to its open source deep learning toolset, DeepSpeed, first released in February. The company says DeepSpeed can help developers train models 15 times larger and 10 times faster using the same computing resources. And they also plan to open source their Turing models so the broader community can build on them.

The general idea is that once one of these very large AI models has been trained, it can actually be customized and employed by other researchers or companies with far fewer resources.

In any case, Microsoft and OpenAI are committed to very large AI, and their new machine may be followed by even bigger systems in the years ahead.

“We’re testing a hypothesis that has been there since the beginning of the field: that a neural network close to the size of the human brain can be trained to be an AGI,” Greg Brockman, OpenAI’s co-founder, chairman, and CTO, told the Financial Times when Microsoft’s investment was first made public. “If the hypothesis is true, the upside for humanity will be remarkable.”

Image credit: Miguel Á. Padriñán / Pexels15

JASON DORRIERJason is managing editor of Singularity Hub. He did research and wrote about finance and economics before moving on to science, technology, and the future. He is curious about pretty much everything, and sad he’ll only ever know a tiny fraction of it all.

Why Am I Still So Angry After All These Years?

Understanding persistent anger and what you can do about it.

Posted May 30, 2020


Has anyone ever said to you, “Why are you still angry about that sexual trauma? Didn’t it happen years ago?” Although this question in itself is discounting (and inaccurate), you may also wonder why anger persists, and whether it will ever go away.

The first part of the answer has to do with the fact that you have a legitimate reason to be angry: You were terribly wronged. This may include being betrayed, lied to, blamed, and violated. Anger is a healthy, appropriate, and reasonable response to an injustice. Sexual trauma is an injustice. It is an illegal, immoral, unethical criminal act, and very wrong.  

Even though it is a criminal act, many, if not most, cases of sexual trauma are not handled well. Many perpetrators are never charged, held accountable, or properly punished, contributing to an unsatisfying lack of closure for victims.

To add fuel to the fire, many, if not most, victims do not feel supported. This includes responses of family, friends, or of professionals in the reporting or treatment process (if reporting or treatment occurred). Victims may find that attempts to talk about it in order to get closure may be met with annoyance, impatience, minimization, and victim-blaming. This only continues the anger and invalidation.

Perpetrators are getting away with committing a crime without consequences, and victims are not supported. Yet, the burden is on the victim to magically let go of the anger?

Part of you knows that it doesn’t feel good to be angry and it would be nice not to be anymore. Part of you may also believe that you need to be angry, because to not be angry would somehow diminish the injustice. As if to say, “Oh well, that’s OK” when it is not OK and never will be. Even if you are working on forgiveness as a path to free yourself, there may be a part of you that doesn’t want the perpetrator to get off the hook.

Other reasons for persistent anger could include a belief that in a way anger protects you by keeping others at a distance, or by creating a shield around you. Or it may even be an odd but familiar companion available to validate your experience when you need it. Anger is also a natural component of grief because it acknowledges what happened. And anger can toggle between other reactions such as shock, disbelief, and depression.

There may also be layers of anger, not only about what happened and how it was handled, but also as a result of the multiple consequences and losses that were incurred because of the trauma. This may include a thwarted career, health issues, consequences in relationships, and difficulty with sexual intimacy. Therefore, there are many reasons why you are still angry after all these years. The events that happened were and are terribly wrong, most likely were not handled well, and lack a fitting closure. Who wouldn’t be angry? It is absolutely justified.

There are also different types of anger. Rage is the most recognizable as an outward display of anger typically loud, big, unpredictable, and dangerous. It is like a large bonfire which can easily get out of control. Everyone knows rage. However, there is another type of anger, not as readily apparent but equally concerning, and that is resentment. Resentment is like the smoldering coals after the flames are gone. This type of anger is held inward. It tends to linger and can churn inside for years. Resentment is caused by unresolved anger. Something happened and there was no acknowledgment, apology, restitution, or closure. Resentment-anger just sits there, and you are left holding a bag of heavy rocks.article continues after advertisement

Sexual trauma is an injustice, but so is persistent anger. Why should your past continue to rob you of your precious time, recalling negative events, feeling toxic emotions of anger, and knowing that it can take a toll on your health and relationships? It is unfair to think you are stuck holding a bag of consequences from anger that could be weighing you down, and zapping your happiness.

How dare the events of the past continue to grip your present and interfere with your future?

Now we have a puzzle. On the one hand, your anger is absolutely valid and justified. On the other hand, you deserve to be happy and free of nagging anger. Understandably, the anger is very deep and well justified. But if you find that anger is creating a barrier to your happiness, then it might be worth considering another path.

Several strategies can help. The first task is to work on a way to get closure. Although you may never get the satisfaction to witness it, if you knew justice would be served, would this help with getting closure? If you knew the perpetrator would carry consequences, would you be willing to wipe your hands of your anger and let the perpetrator hold the bag of rocks instead of you?

I offer the concept of poetic justice. This is the concept that somehow, somewhere, someway, people reap the natural consequences of their own behavior. What the perpetrator did was intentional. This person planned, prepared, and knowingly took something that was not given. People who act with malicious malintent toward others will invariably have consequences…even if they lie and seem to get away with the crime. That person knows what was done, you know what was done, and others also probably know. Somehow life has a way of finding justice. These consequences are not only inevitable, but it is that person’s journey, his or her life lessons. You no longer need to be stuck in anger-jail waiting for the other person to get punished. You can be free knowing that somehow, somewhere, some way, justice will be served. You don’t have to do anything. They will do it to themselves.article continues after advertisement

Your challenge is to release your bond with this person that hurt you. Poetic justice gives you a way to release your anger without minimizing what happened. It gives you permission to free yourself of the past, and from constant thoughts about what happened. This makes room to think about new things, and for new things to come into your future.

Visualization Exercise: With practice, anger can subside, release, and fade away.  Imagine a theatrical stage in a large venue. See the character Anger dominating center stage of your life, taking a disproportionate amount of your attention and energy. You are the director. Thank Anger for its performance and imagine guiding it to the side of the stage. Now imagine Anger going further and further away. Now Anger is sitting in a single seat at the back of the theater. It is not even on stage anymore, but rather a passive member of the audience…Notice that there is room for something new to come on stage. Imagine something wonderful on your stage.

A Strange Thing Can Happen When You Use Mouthwash After You Exercise


In ads on TV, it all looks so simple. People use mouthwash, it instantly neutralises all the nasty bacteria hiding in their mouths, and – just like that – their dental hygiene is assured.

But what’s really going on when you rinse a cap-load of antibacterial chemicals around your mouth? What does that to your body, and to other kinds of microorganisms that may actually be beneficial to health?

As a study showed last year, the downstream effects can be surprising, and far-reaching too, affecting much more than just your dental wellbeing.

In an experiment led by scientists from the UK and Spain, researchers found that the simple act of using mouthwash after exercising can reduce one of the benefits of exercise: lowering blood pressure.

When you exercise, your blood vessels open in response to the production of nitric oxide, which increases the diameter of blood vessels. This process is called vasodilation, and it increases blood flow circulation to active muscles.

For a long time, researchers thought this only happened during exercise, but in more recent years, evidence has shown that circulation stays high (meaning blood pressure is lowered) even after exercise – thanks to how bacteria interact with a compound called nitrate, which forms when nitric oxide degrades.

“Research over the last decade has shown that nitrate can be absorbed in the salivary glands and excreted with saliva in the mouth,” explains physiology specialist Raul Bescos from Plymouth University.

“Some species of bacteria in the mouth can use nitrate and convert into nitrite – a very important molecule that can enhance the production of nitric oxide in the body.”

Once nitrite is produced and swallowed with saliva, it becomes absorbed into blood circulation and reduces back to nitric oxide, which keeps blood vessels wide and lowers blood pressure.

But according to this small study, it looks like this biological mechanism can be significantly interrupted if anti-bacterial mouthwash gets added into the post-exercise mix.

In an experiment, 23 healthy adults ran on a treadmill for 30 minutes. After the workout, the participants were asked to rinse their mouth with either an antibacterial mouthwash or a mint-flavoured placebo.

These mouth rinses occurred immediately after the exercise, and also at 30, 60, and 90 minutes after.

The participants also had their blood pressure taken during the experiment, immediately after the exercise, and during their rest period.

The results showed that at one hour after the treadmill session, average reduction in systolic blood pressure in the placebo group was –5.2 mmHg (millimetres of mercury).

The reduction in the mouthwash-using group was much lower, showing an average of –2.0 mmHg at the same point in time, suggesting the use of the antibacterial mouthwash (0.2 percent chlorhexidine) had lowered the systolic blood pressure reduction by over 60 percent.

At the end of the monitoring window, two hours after the treadmill session, the mouthwash group showed no sign of blood pressure reduction stemming from the exercise, whereas the placebo group still showed a significant reduction compared to their pre-exercise values.

“This is the first evidence showing that the nitrate-reducing activity of oral bacteria is a key mechanism to induce the acute cardiovascular response to exercise during the recovery period in healthy individuals,” the authors explained in their paper.

While it’s only a small study, it serves as an important reminder of how not all bacteria are necessarily bad for us – and that ingesting antibacterial chemicals that indiscriminately terminate mouth-dwelling microbes could hamper important biological processes necessary for good health.

“These findings show that nitrite synthesis by oral bacteria is hugely important in kick-starting how our bodies react to exercise over the first period of recovery, promoting lower blood pressure and greater muscle oxygenation,” said one of the team, nutritionist Craig Cutler.

“In effect, it’s like oral bacteria are the ‘key’ to opening up the blood vessels. If they are removed, nitrite can’t be produced and the vessels remain in their current state.”

The findings have been reported in Free Radical Biology and Medicine.

A version of this article was first published in September 2019.

Learn More

  1. Researchers report impact of diet on incidence of migraineHealio
  2. Study shows eating vegetables rich in nitrates can lower blood pressureHealio
  1. Is this relapsing remitting MS? How do you know and what will you do?Patricia Coyle, MD, myCME, 2019
  2. Coronavirus Dx Emergency Use Authorizations Progressing Rapidly Despite CriticismMadeleine Johnson, 360Dx, 2020

New Recombineering Method May Overcome Key Obstacle in Genetically Engineering Bacteria

May 29, 20200

Colorized scanning electron micrograph, SEM, of Pseudomonas aeruginosa bacteria. Callista Images/Getty Images Share

SEM of Pseudomonas aeruginosa bacteria

May 29, 20200

Colorized scanning electron micrograph, SEM, of Pseudomonas aeruginosa bacteria. Callista Images/Getty Images

Researchers in the Blavatnik Institute at Harvard Medical School (HMS) and the Biological Research Center in Szeged, Hungary, say they have developed a new genetic engineering technique that promises to transform recombineering methodology.

Their report “Improved bacterial recombineering by parallelized protein discovery” is published in PNAS.

The team created a high-throughput screening technology to look for the most efficient proteins that serve as the engines of recombineering. Such single-stranded DNA-annealing proteins (SSAPs) reside within phages.

Applying the new method, which enables the screening of more than two hundred SSAPs, the researchers identified two proteins that appear to be particularly promising. One of them doubled the efficiency of single-spot edits of the bacterial genome. It also improved tenfold the ability to perform multiplex editing—making multiple edits genome-wide at the same time.

The other one enabled recombineering in the human pathogen Pseudomonas aeruginosa, a frequent cause of life-threatening, hospital-acquired infections, for which there has long been a dearth of good genetic tools.

“Exploiting bacteriophage-derived homologous recombination processes has enabled precise, multiplex editing of microbial genomes and the construction of billions of customized genetic variants in a single day, according to the scientists. The techniques that enable this, multiplex automated genome engineering (MAGE) and directed evolution with random genomic mutations (DIvERGE), are however, currently limited to a handful of microorganisms for which…SSAPs… that promote efficient recombineering have been identified,” write the investigators.

“Thus, to enable genome-scale engineering in new hosts, efficient SSAPs must first be found. Here we introduce a high-throughput method for SSAP discovery that we call serial enrichment for efficient recombineering’ (SEER). By performing SEER in Escherichia coli to screen hundreds of putative SSAPs, we identify highly active variants PapRecT and CspRecT, which increases the efficiency of single-locus editing to as high as 50% and improves multiplex editing by 5- to 10-fold in E. coli, while PapRecT enables efficient recombineering in Pseudomonas aeruginosa.”

“CspRecT and PapRecT are also active in other, clinically and biotechnologically relevant enterobacteria. We envision that the deployment of SEER in new species will pave the way toward pooled interrogation of genotype-to-phenotype relationships in previously intractable bacteria.”

“Recombineering will be a very critical tool that will augment our DNA writing and editing capabilities in the future, and this is an important step in improving the efficiency and reach of the technology,” said study first author Timothy Wannier, PhD, research associate in genetics in lab of George Church, the Robert Winthrop Professor of Genetics at HMS.

Previous genetic engineering methods, including CRISPR Cas9-based gene-editing, have been ill-suited to bacteria because these methods involve “cutting and pasting” DNA, the researchers said. This is because, unlike multicellular organisms, bacteria lack the machinery to repair double-stranded DNA breaks efficiently and precisely, thus DNA cutting can profoundly interfere with the stability of the bacterial genome, Wannier pointed out.

The advantage of recombineering is that it works without cutting DNA. Instead, recombineering involves sneaking edits into the genome during bacterial reproduction. Bacteria reproduce by splitting in two. During that process, one strand of their double-stranded, circular DNA chromosomes goes to each daughter cell, along with a new second strand that grows during the early stages of fission.

The raw materials for recombineering are short, approximately 90 base strands of DNA that are made to order. Each strand is identical to a sequence in the genome, except for edits in the strand’s center. These short strands slip into place as the second strands of the daughter cells grow, efficiently incorporating the edits into their genomes.

Among many possible uses, edits might be designed to interfere with a gene in order to pinpoint its function or, alternatively, to improve production of a valuable bacterial product. SSAPs mediate attachment and proper placement of the short strand within the growing new half of the daughter chromosome.

Recombineering might enable the substitution of a naturally occurring bacterial amino acid with an artificial one. Among other things, doing so could enable the use of bacteria for environmental cleanup of oil spills or other contaminants, that depend on these artificial amino acids to survive, meaning that the modified bacteria could be easily annihilated once the work is done to avoid the risks of releasing engineered microbes into the environment, Wannier said.

“The bacteria would require artificial amino acid supplements to survive, meaning that they are preprogrammed to perish without the artificial feed stock,” Wannier added.

A version of recombineering, called multiplex automated genome engineering (MAGE), could greatly boost the benefits of the technique. The particular advantage of MAGE is its ability to make multiple edits throughout the genome in one fell swoop.

MAGE could lead to progress in projects requiring reengineering of entire metabolic pathways, said John Aach, PhD, lecturer in genetics at HMS. Case in point, Aach added, are large-scale attempts to engineer microbes to turn wood waste into liquid fuels.

“Many investigator-years’ effort in that quest have made great progress, even if they have not yet produced market-competitive products,” he said, noting that such endeavors require testing many combinations of edits. “We have found that using MAGE with a library of DNA sequences is a very good way of finding the combinations that optimize pathways.”

A more recent descendant of recombineering, named directed evolution with random genomic mutations (DIvERGE), promises benefits in the fight against infectious diseases and could open new avenues for tackling antibiotic resistance.

By introducing random mutations into the genome, DIvERGE can speed up natural bacterial evolution. This helps researchers quickly uncover changes that could arise naturally in harmful bacteria that would make them resistant to antibiotic treatment, explained Akos Nyerges, PhD, research fellow in genetics in Church’s lab at HMS, previously at the Biological Research Center of the Hungarian Academy of Sciences.

“Improvements in recombineering will allow researchers to more quickly test how bacterial populations can gain resistance to new antibacterial drugs, helping researchers to identify less resistance-prone antibiotics,” Nyerges said.

Recombineering will likely usher in a whole new world of applications that would be hard to foresee at this juncture, according to the researchers.

“The new method greatly improves our ability to modify bacteria,” Wannier said. “If we could modify a letter here and there in the past, the new approach is akin to editing words all over a book and doing so opens up the scientific imagination in a way that was not previously possible.”