Select Page
Bird Flu Is Spreading in Alarming New Ways

Bird Flu Is Spreading in Alarming New Ways

As a recent example of what may ensue, Pitesky points to the repeated African swine fever outbreaks across various Asian countries in the past decade, which decimated the pig farming industry to the extent that pork was briefly usurped by poultry as the most widely consumed animal protein on the planet. Pitesky argues, however, that the current model of governments heavily compensating farmers for their livestock losses in the wake of a viral outbreak is financially unsustainable, and more investment needs to be diverted toward AI-driven technologies that can prevent these infections in the first place.

“I work on predictive models, using a combination of weather radar, satellite imagery, and machine learning, to understand how waterfowl behavior around different farms is changing,” says Pitesky. “We can use this information to understand which of the 50,000 to 60,000 commercial poultry facilities in the US are at most risk, and form strategies to protect all the birds in those facilities.”

Technology may ultimately offer a path toward eliminating the virus in commercial poultry. In October, a team of researchers in the UK published a study in the journal Nature Communications demonstrating that it is possible to use the gene-editing tool Crispr to make chickens resistant to avian influenza. This was done through editing genes that make the proteins ANP32A, ANP32B, and ANP32E in chickens, which the virus uses to gain access to chicken cells.

Crispr has been shown to be capable of making livestock resistant to other infections such as the cancer-causing viral disease avian leukosis and porcine reproductive and respiratory syndrome, which is responsible for widespread economic losses in pig farms.

“The currently available methods are the use of strict farm biosecurity, poultry vaccinations in some countries, and massive depopulation of infected or exposed chicken flocks,” says Alewo Idoko-Akoh at the University of Bristol, the lead researcher on the Nature Communications study. “These methods have been partially successful but have so far failed to stop recurrent bird flu outbreaks around the world. Gene editing of chickens to introduce disease resistance should be considered as an additional tool for preventing or limiting the spread of bird flu.”

Pitesky described the paper as “really interesting” but pointed out that it would require widespread public acceptance toward consuming gene-edited chicken for it to become commercially viable. “I think that those technological solutions have a lot of potential, but the issue more than anything, especially in the United States, is sentiment toward chickens that have been genetically modified,” he says.

For now, Iqbal says that the best chance of keeping avian influenza under control is more active surveillance efforts in animal populations around the world, to understand how and where the H5N1 is spreading.

“The surveillance system has been improved, and any infection that appears unusual is thoroughly investigated,” he says of the situation in the US. “This has helped to identify unusual outbreaks, such as infections in goats and cattle.” However, he says, much more work is needed to detect the virus in animals that don’t show signs of disease.

The World’s E-Waste Has Reached a Crisis Point

The World’s E-Waste Has Reached a Crisis Point

The phone or computer you’re reading this on may not be long for this world. Maybe you’ll drop it in water, or your dog will make a chew toy of it, or it’ll reach obsolescence. If you can’t repair it and have to discard it, the device will become e-waste, joining an alarmingly large mountain of defunct TVs, refrigerators, washing machines, cameras, routers, electric toothbrushes, headphones. This is “electrical and electronic equipment,” aka EEE—anything with a plug or battery. It’s increasingly out of control.

As economies develop and the consumerist lifestyle spreads around the world, e-waste has turned into a full-blown environmental crisis. People living in high-income countries own, on average, 109 EEE devices per capita, while those in low-income nations have just four. A new UN report finds that in 2022, humanity churned out 137 billion pounds of e-waste—more than 17 pounds for every person on Earth—and recycled less than a quarter of it.

That also represents about $62 billion worth of recoverable materials, like iron, copper, and gold, hitting e-waste landfills each year. At this pace, e-waste will grow by 33 percent by 2030, while the recycling rate could decline to 20 percent. (You can see this growth in the graph below: purple is EEE on the market, black is e-waste, and green is what gets recycled.)

Graph displaying ewaste generation

Courtesy of UN Global E-waste Statistics Partnership

“What was really alarming to me is that the speed at which this is growing is much quicker than the speed that e-waste is properly collected and recycled,” says Kees Baldé, a senior scientific specialist at the United Nations Institute for Training and Research and lead author of the report. “We just consume way too much and we dispose of things way too quickly. We buy things that we may not even need, because it’s just very cheap. And also these products are not designed to be repaired.”

Humanity has to quickly bump up those recycling rates, the report stresses. In the first pie chart below, you can see the significant amount of metals we could be saving, mostly iron (chemical symbol Fe, in light gray), along with aluminum (Al, in dark gray), copper (Cu), and nickel (Ni). Other EEE metals include zinc, tin, and antimony. Overall, the report found that in 2022, generated e-waste contained 68 billion pounds of metal.

Graphs displaying recoverable and nonrecoverable metals in ewaste

Courtesy of UN Global E-waste Statistics Partnership

Never-Repeating Patterns of Tiles Can Safeguard Quantum Information

Never-Repeating Patterns of Tiles Can Safeguard Quantum Information

This extreme fragility might make quantum computing sound hopeless. But in 1995, the applied mathematician Peter Shor discovered a clever way to store quantum information. His encoding had two key properties. First, it could tolerate errors that only affected individual qubits. Second, it came with a procedure for correcting errors as they occurred, preventing them from piling up and derailing a computation. Shor’s discovery was the first example of a quantum error-correcting code, and its two key properties are the defining features of all such codes.

The first property stems from a simple principle: Secret information is less vulnerable when it’s divided up. Spy networks employ a similar strategy. Each spy knows very little about the network as a whole, so the organization remains safe even if any individual is captured. But quantum error-correcting codes take this logic to the extreme. In a quantum spy network, no single spy would know anything at all, yet together they’d know a lot.

Each quantum error-correcting code is a specific recipe for distributing quantum information across many qubits in a collective superposition state. This procedure effectively transforms a cluster of physical qubits into a single virtual qubit. Repeat the process many times with a large array of qubits, and you’ll get many virtual qubits that you can use to perform computations.

The physical qubits that make up each virtual qubit are like those oblivious quantum spies. Measure any one of them and you’ll learn nothing about the state of the virtual qubit it’s a part of—a property called local indistinguishability. Since each physical qubit encodes no information, errors in single qubits won’t ruin a computation. The information that matters is somehow everywhere, yet nowhere in particular.

“You can’t pin it down to any individual qubit,” Cubitt said.

All quantum error-correcting codes can absorb at least one error without any effect on the encoded information, but they will all eventually succumb as errors accumulate. That’s where the second property of quantum error-correcting codes kicks in—the actual error correction. This is closely related to local indistinguishability: Because errors in individual qubits don’t destroy any information, it’s always possible to reverse any error using established procedures specific to each code.

Taken for a Ride

Zhi Li, a postdoc at the Perimeter Institute for Theoretical Physics in Waterloo, Canada, was well versed in the theory of quantum error correction. But the subject was far from his mind when he struck up a conversation with his colleague Latham Boyle. It was the fall of 2022, and the two physicists were on an evening shuttle from Waterloo to Toronto. Boyle, an expert in aperiodic tilings who lived in Toronto at the time and is now at the University of Edinburgh, was a familiar face on those shuttle rides, which often got stuck in heavy traffic.

“Normally they could be very miserable,” Boyle said. “This was like the greatest one of all time.”

Before that fateful evening, Li and Boyle knew of each other’s work, but their research areas didn’t directly overlap, and they’d never had a one-on-one conversation. But like countless researchers in unrelated fields, Li was curious about aperiodic tilings. “It’s very hard to be not interested,” he said.

Selective Forgetting Can Help AI Learn Better

Selective Forgetting Can Help AI Learn Better

The original version of this story appeared in Quanta Magazine.

A team of computer scientists has created a nimbler, more flexible type of machine learning model. The trick: It must periodically forget what it knows. And while this new approach won’t displace the huge models that undergird the biggest apps, it could reveal more about how these programs understand language.

The new research marks “a significant advance in the field,” said Jea Kwon, an AI engineer at the Institute for Basic Science in South Korea.

The AI language engines in use today are mostly powered by artificial neural networks. Each “neuron” in the network is a mathematical function that receives signals from other such neurons, runs some calculations, and sends signals on through multiple layers of neurons. Initially the flow of information is more or less random, but through training, the information flow between neurons improves as the network adapts to the training data. If an AI researcher wants to create a bilingual model, for example, she would train the model with a big pile of text from both languages, which would adjust the connections between neurons in such a way as to relate the text in one language with equivalent words in the other.

But this training process takes a lot of computing power. If the model doesn’t work very well, or if the user’s needs change later on, it’s hard to adapt it. “Say you have a model that has 100 languages, but imagine that one language you want is not covered,” said Mikel Artetxe, a coauthor of the new research and founder of the AI startup Reka. “You could start over from scratch, but it’s not ideal.”

Artetxe and his colleagues have tried to circumvent these limitations. A few years ago, Artetxe and others trained a neural network in one language, then erased what it knew about the building blocks of words, called tokens. These are stored in the first layer of the neural network, called the embedding layer. They left all the other layers of the model alone. After erasing the tokens of the first language, they retrained the model on the second language, which filled the embedding layer with new tokens from that language.

Even though the model contained mismatched information, the retraining worked: The model could learn and process the new language. The researchers surmised that while the embedding layer stored information specific to the words used in the language, the deeper levels of the network stored more abstract information about the concepts behind human languages, which then helped the model learn the second language.

“We live in the same world. We conceptualize the same things with different words” in different languages, said Yihong Chen, the lead author of the recent paper. “That’s why you have this same high-level reasoning in the model. An apple is something sweet and juicy, instead of just a word.”

Solar-Powered Farming Is Quickly Depleting the World’s Groundwater Supply

Solar-Powered Farming Is Quickly Depleting the World’s Groundwater Supply

That is certainly the case in Yemen, on the south flank of the Arabian Peninsula, where the desert sands have a new look these days. Satellite images show around 100,000 solar panels glinting in the sun, surrounded by green fields. Hooked to water pumps, the panels provide free energy for farmers to pump out ancient underground water. They are irrigating crops of khat, a shrub whose narcotic leaves are the country’s stimulant of choice, chewed through the day by millions of men.

For these farmers, the solar irrigation revolution in Yemen is born of necessity. Most crops will only grow if irrigated, and the country’s long civil war has crashed the country’s electricity grid and made supplies of diesel fuel for pumps expensive and unreliable. So, they are turning en masse to solar power to keep the khat coming.

The panels have proved an instant hit, says Middle East development researcher Helen Lackner of SOAS University of London. Everybody wants one. But in the hydrological free-for-all, the region’s underground water, a legacy of wetter times, is running out.

The solar-powered farms are pumping so hard that they have triggered “a significant drop in groundwater since 2018 … in spite of above average rainfall,” according to an analysis by Leonie Nimmo, a researcher who was until recently at the UK-based Conflict and Environment Observatory. The spread of solar power in Yemen “has become an essential and life-saving source of power,” both to irrigate food crops and provide income from selling khat, he says, but it is also “rapidly exhausting the country’s scarce groundwater reserves.”

In the central Sana’a Basin, Yemen’s agricultural heartland, more than 30 percent of farmers use solar pumps. In a report with Musaed Aklan, a water researcher at the Sana’a Center for Strategic Studies, Lackner predicts a “complete shift” to solar by 2028. But the basin may be down to its last few years of extractable water. Farmers who once found water at depths of 100 feet or less are now pumping from 1,300 feet or more.

Some 1,500 miles to the northeast, in in the desert province of Helmand in Afghanistan, more than 60,000 opium farmers have in the past few years given up on malfunctioning state irrigation canals and switched to tapping underground water using solar water pumps. As a consequence, water tables have been falling typically by 10 feet per year, according to David Mansfield, an expert on the country’s opium industry from the London School of Economics.

An abrupt ban on opium production imposed by Afghanistan’s Taliban rulers in 2022 may offer a partial reprieve. But the wheat that the farmers are growing as a replacement is also a thirsty crop. So, water bankruptcy in Helmand may only be delayed.

“Very little is known about the aquifer [in Helmand], its recharge or when and if it might run dry,” according to Mansfield. But if their pumps run dry, many of the million-plus people in the desert province could be left destitute, as this vital desert resource—the legacy of rainfall in wetter times—disappears for good.