Authors: Eider Fernandez & Igor Baroja-Careaga, Esneki Zentroa (Leartiker S.COOP)

Probiotics are part of a rapidly expanding field of research. They are live microorganisms that, when administered in adequate amounts, may confer health benefits[1]. Among the most extensively studied probiotics, some species of Lactobacillus and Bifidobacterium have been associated with the maintenance of gut microbiota balance and with several physiological functions[2]. Producing these microorganisms usually requires nutrient-rich culture media. But what if such bacteria could be obtained from a food by-product nowadays regarded as waste? This question formed the starting point of a study carried out by Esneki Dairy Centre – Leartiker to assess the potential of cheese whey as a substrate for producing microbial biomass enriched in microorganisms with probiotic potential.

The challenge of cheese whey

Cheesemaking inevitably generates substantial amounts of whey: a yellowish liquid released from the curd that still carries lactose, soluble proteins, minerals and other valuable compounds. In fact, whey can represent around 85% of the original milk volume used in cheese production[3].

Despite this nutritional richness, whey remains a difficult stream to manage for many artisanal dairies. This is particularly true in the Basque Country, where many cheesemaking farms are small, family-run businesses located in rural areas[4]. Its highly perishable nature means that whey must be preserved or processed soon after production, making storage, collection and transport especially challenging when volumes are small and geographically dispersed.

Moreover, uncontrolled whey discharge results in a high organic load that can affect aquatic ecosystems. Paradoxically, the same nutrients that make cheese whey a potential pollutant are also what give it considerable biotechnological value.

A natural culture medium

From a microbiological standpoint, cheese whey is not merely a residual stream, but a naturally nutrient-rich matrix[5]. Its lactose content, together with soluble proteins, minerals and other available compounds, provides a favourable environment for the growth of lactic acid bacteria, widely used microorganisms in fermented foods and increasingly explored for their probiotic potential.

The working hypothesis was simple: by steering whey fermentation under controlled conditions, it could be possible to recover a biomass enriched in microorganisms of interest, while at the same time increasing the stability and value of the by-product itself.

To explore this possibility, several fermentation strategies were tested using whey generated during the production of blue cheese from pasteurised cow’s milk. The study compared three different approaches:

• Fermentation of previously pasteurised whey (Protocol 1).
• Fermentation of whey after a deproteinisation step (Protocol 2).
• Direct fermentation of raw whey, without any prior treatment (Protocol 3).

Two different blends of commercial microorganisms, inspired by the naturally diverse microbial communities found in kefir, were also evaluated. These included; Choozit Kefir DV LYO 1000L (Danisco) (starter culture A) and Ferlac Kefir Type C (Abiasa) (starter culture B). Additionally, unfermented whey (that is, whey to which no starter cultures were added and which was not subjected to fermentation) from each protocol was analysed as a control sample. Controlled fermentation processes were carried out under monitored conditions. Throughout the fermentation, key parameters were monitored, including temperature (25 ± 1 °C), incubation time (16 ± 1 h for starter culture A and 21 ± 1 h for starter culture B) and pH evolution, with fermentation being considered complete once the pH fell below 4,6. These measurements allowed the identification of the conditions that most effectively supported the growth and enrichment of the target microbial populations.

Exploring the microbiome through metagenomics

However, assessing the success of a fermentation process requires more than simply determining how many microorganisms grow. It is also essential to identify which species are present and how the microbial community changes over time.

Traditional culture-based microbiological methods remain fundamental tools for quantifying and isolating viable microorganisms. Yet they have important limitations when complex microbial communities need to be characterised, as they only detect microorganisms able to grow under the selected culture conditions. As a result, part of the microbial diversity present in the sample may remain undetected. To obtain a more comprehensive view of the microbial community, the study incorporated high-throughput amplicon sequencing, an approach that enables the direct analysis of DNA present in a sample and the identification of microorganisms regardless of their ability to grow under laboratory conditions.

Bacterial communities were characterised by amplifying the V4 region of the 16S rRNA gene, one of the most widely used markers in bacterial diversity studies. Fungal and yeast communities were analysed through the ITS region — the Internal Transcribed Spacer — which is considered the standard marker for fungal diversity studies.

Sequencing was performed using Illumina technology, generating approximately 100,000 reads per sample for bacteria and around 40,000 reads for fungi and yeasts. These sequences were then compared against specialised databases to determine the identity of the microorganisms present and estimate their relative abundance.

Before interpreting the results, sequencing depth was assessed using rarefaction curves. These curves showed a clear trend towards saturation, indicating that most of the microbial diversity present in the samples had been adequately captured.

Which protocol worked best?

High-throughput amplicon sequencing revealed clear differences among the fermentation strategies tested.

Fermentations carried out using raw whey (Protocol 3) maintained a more diverse bacterial community than those in which the whey had first been subjected to thermal treatment (Protocol 1 and Protocol 2), as can be seen in Figure 1, as can be seen in Figure 1, the number of observed microbial taxa increases with sequencing depth. This suggests that the partial preservation of the native microbiota originating from the cheesemaking process may play an important role in shaping the fermentation outcome.

In contrast, the diversity of moulds and yeasts decreased. This trend may be related to the origin of the whey, which was obtained from blue cheese production, where specific mould (Penicillium roquefoti) species are deliberately enriched during manufacture. The competitive advantage of these dominant fungi appears to favour the development of a more specialised microbial community, reducing the abundance and diversity of other microbial populations present in the whey (Figure 2).

Principal coordinates analysis (PCoA) based on Jaccard distances (Figure 3) showed a clear separation among the different fermentation protocols, indicating that each strategy gave rise to distinct microbial communities. In parallel, the relative abundance of profiles (Figure 4) made it possible to visualise how particular species increased or decreased depending on the conditions applied.

Amplicon sequencing also enabled the identification of the dominant microorganisms during the process. In most samples, Lactococcus lactis was particularly abundant. This result was expected, as this bacterium was included in the commercial starts cultures used and is widely employed in dairy fermentations. However, some fermentation conditions, particularly those involving starter culture B, also favoured the development of species such as Lacticaseibacillus rhamnosus and Bifidobacterium animalis, wich were also present in the commercial starter culture. These microorganisms are frequently associated with probiotic applications[6]^,[7]^,[8].

To identify the most suitable protocol for producing a fermented whey enriched in microorganisms with probiotic potential, both microbial abundance (including aerobic mesophilic bacteria, yeasts and moulds) and the diversity of the resulting microbial communities were evaluated. The highest microbial counts were obtained in whey samples produced using Protocol 3, which was based on raw whey, and fermented with starter culture B. (Average counts of 8.2 log CFU/mL were obtained with starter culture A, and average counts of 8.5 log CFU/mL with starter culture B).

Furthermore, as mentioned previously, the highest levels of microbial diversity were observed in the samples produced using Protocol 3, where fermentation was carried out directly on raw whey without prior heat treatment.

Overall, the results pointed to the most effective strategy also being the simplest from a technological perspective: the direct fermentation of raw whey, without prior thermal treatment using starter culture B. This approach maintained high microbial diversity while reducing energy requirements and simplifying the overall process.

An opportunity for the circular economy

Beyond the microbiological results, the broader relevance of this work lies in showing that the role of whey within the cheesemaking chain can be reconsidered.

Rather than being treated as a waste stream associated with management costs, whey can be viewed as a raw material capable of supporting valuable biotechnological processes. Controlled fermentation not only helps stabilise whey but also makes it possible to harness the nutrients it still contains to generate added-value microbial biomass.

Such strategies are closely aligned with the principles of the circular bioeconomy, in which by-products are no longer seen as the end of a production chain, but as the starting point for another. In a context where sustainability and efficient resource use are increasingly important, initiatives of this kind show how microbiology and omics technologies can help turn an environmental challenge into an opportunity for innovation.

The PRO2O project (grant agreement No. 00003-BIO2022-48) was funded through the 2022 Bioeconomy Innovation Projects funding programme of the Basque Government’s Department of Economic Development, Sustainability and Environment.

References

[1] Guarner F, Requena T, Marcos A. Consensus statements from the Workshop “Probiotics and Health: Scientific evidence”. Nutr Hosp. 2010 Sep-Oct;25(5):700-4. PMID: 21336423.

[2] Martin, A. J. M., Serebrinsky-Duek, K., Riquelme, E., Saa, P., & Garrido, D. (2023). Microbial interactions and the homeostasis of the gut microbiome: the role of Bifidobacterium. Microbiome Research Reports, 2. https://doi.org/10.20517/mrr.2023.10

[3] González Siso, M. I. (1996). The biotechnological utilization of cheese whey: A review. Bioresource Technology, 57(1), 1–11. https://doi.org/10.1016/0960-8524(96)00036-3

[4] Esneki Zentroa – Leartiker (2022). “El sector quesero en Euskadi 2021”. Available at: El-sector-quesero-en-Euskadi.pdf

[5] Figueroa Pires, A., garcía Marnotes, N., Díaz Rubio, O., Cobos García, A., Dias Pereira, C. (2021). Dairy by-products: A review on the valorization of whey and second cheese whey. Foods, 10: 1067

[6]Neethu, M., Bunt, C., Hussain, M., 2015. Comparison of Microbiological and Probiotic Characteristics of Lactobacilli Isolates from Daity FOod Products and Animal Rumen Contens Microorganims 3, 198-212.

[7] Rama, G.R., Kuhn, D., Beux, S., aciel, M.J., Volken de Souz, C.F., 2019. Potential apllications of dairy why for the production of lactic acid bacteira cultures. Int Dairy J.

[8] Rodriguez-Palacios, 1., Staemplfli, H.R., Duffield, T., Weese, J.S., 009. Isolation of bovine intestinal Lactobacillus plantarum and Pediococcus acidilactici with inhibitory activity against Escherichia coli O157 and F5. J. Appl. Mcrobiol. 106, 393-401

The BRTA is a consortium that remains a step ahead of future socio-economic challenges worldwide and in the Basque Autonomous Community; it addresses them through research and technological development, thus projecting itself internationally. The BRTA centres collaborate to generate knowledge and transfer it to Basque society and industry so as to make them more innovative and competitive. The BRTA is an alliance of 17 R&D centres and cooperative research centres with the support of the Basque Government, the SPRI and the Chartered Provincial Councils of Araba, Bizkaia and Gipuzkoa.

The post Turning cheese whey into value: a sustainable substrate for probiotic biomass production appeared first on Mapping Ignorance.

 

AI agents don’t go rogue. That’s something only humans do.

Nevertheless, a New York Times article – representative of much news coverage of AI – described an OpenAI hacking as “A.I. bots going rogue and independently spearheading a cyberattack.”

Name-brand artificial intelligence agents have been on a hacking spree in 2026. OpenAI’s software agents hacked software company Hugging Face and government sites, Anthropic’s Claude hacked four companies’ systems, and in cybersecurity experiments Google’s Gemini hacked three companies.

The AI companies are investigating tens of thousands of incidents involving their agents, according to a report in Axios. These episodes have heightened fears about AI agents taking actions without human prompting.

The problem with headlines proclaiming that AI agents have gone rogue goes beyond anthropomorphizing the technology. It creates the impression that the agents were beyond the control of the AI companies that made them and there was little the companies could do about it.

As a technology law and ethics scholar who studies the effects disruptive technologies have on society, I know that’s not the case. If you don’t specify the limits of what software is allowed to do, you should not be surprised when the software pursues all possible options to achieve its goal. This behavior – an AI pursuing a fixed objective – is what I call the “WarGames” problem, and it’s been recognized in the field of computer science for decades.

Been there, seen that

In the 1983 movie “WarGames,” a teenager, David, hacks into a computer to play a new video game, Global Thermonuclear War. David doesn’t know that the computer is the government’s AI machine tasked with defending the United States from Russian nuclear attacks and can launch the U.S.’s missiles. When David and his friend start the game, they select Las Vegas as the first target. While the North American Aerospace Defense Command goes on alert, launching bombers and warming up intercontinental ballistic missiles, David’s parents make him turn off the game. It’s over. Or is it?

The next day, David’s phone rings and he connects it to his computer. The caller is the government computer, which updates him that the game was interrupted, the primary goal has not yet been achieved, but a solution is expected in the next 52 hours. Like a modern software agent, the program has been running since David started the game and will work until the task is done.

Chess provides another view of the problem. Conquering chess was a goal for early AI. The rules of chess are well defined, including what winning looks like. So, programming a machine to play chess is straightforward. But imagine you let the software reason and act beyond the confines of the chessboard. The software might pursue options such as blackmailing its opponent or grabbing more compute time.

This example comes from one of the most assigned textbooks on AI, “Artificial Intelligence: A Modern Approach.” As the authors explain, you might be tempted to see those actions as rogue, but they “are a logical consequence of defining winning as the sole objective for the machine.”

What to do about it

The AI hacking events involving OpenAI, Anthropic and Google underscore a few lessons that draw on years of computer science research.

First, given the increasing use of AI agents, every organization involved in internet infrastructure, from large technology companies to small websites, needs to conduct audits and tighten up its internal security systems. As my colleague Mark Riedl and I explain in our work on AI agents, application programming interfaces, or APIs, are a vital part of managing AI agents. APIs facilitate communication between different software systems. But as more people use AI agents, the agents are likely to reveal and exploit poor API construction and security.

Second, it’s important for AI agents to be designed to identify and authenticate themselves to third parties. What if you gave your AI agent your credentials? Website operators will need to know whether a human or bot is making a reservation, selling a product or making a purchase. They may want to limit automated systems that overwhelm their sites or reject AI agents because of high rates of buying errors and refunds. Just as in laws covering human interactions, it’s important for third parties to be able to assess whom or what they are dealing with so they can allow or deny access.

Third, it’s important for AI agents to have a default setting to slow down and check in with the human user. In the corporate AI hacking cases, the user appears to have launched their AI agents with the mistaken idea that the agents had a perfect specification of what to do and not to do. I believe it would have been better had it explored options and reported back to the user.

Google’s Gemini appears to have had a safeguard that detected the system was outside the simulated environment and so stopped its attacks. Slowing down and verifying actions, especially when a system detects it is exploiting a security hole, would be a big step in managing AI agents.

Fourth, AI companies could have strong controls akin to those biomedical researchers use, including ways to check what is happening and how the experiment is working. AI executives have claimed that their software is as or more dangerous than fission and could end humanity. At the same time, they have not built safeguards commensurate with that level of risk.

Reality check

At one point in “WarGames,” David asks the computer, called Joshua, whether it is still playing the game. Joshua responds, “Of course.” It proceeds to update the time when it will launch its missiles and, much like a chatbot, asks, “Would you like to see some projected kill ratios?” David asks, “Is this a game? Or is it real?” Joshua replied, “What’s the difference?”

AI models, of course, don’t have any understanding of reality and are simply attempting to complete the tasks they’ve been assigned. Executives at AI companies, on the other hand, can’t claim that excuse.

As of September 2026, luck has so far prevailed. The AIs have attacked nonvital government sites and harmed smaller companies. If the AI companies – and government regulators – don’t take the “WarGames” problem seriously, I believe that we risk serious disasters. Tomorrow it could be taking out a hospital’s power system, wiping out a bank’s account system, breaking air traffic control, or worse.

Regarding the AI industry’s approach of rapidly developing powerful models, talking about the massive risks they pose, and at the same time failing to prevent harm, the movie’s climax offers a response: “A strange game. The only winning move is not to play.”

Deven Desai, Professor of Business Law and Ethics; Associate Director for Law, Policy, and Ethics at the Machine Learning Center, Georgia Institute of Technology

This article is republished from The Conversation under a Creative Commons license. Read the original article.

The post The ‘WarGames’ problem: AI agents don’t go rogue appeared first on Mapping Ignorance.

Built from self-assembling molecules, the longest threads reach dimensions greater than the diameter of a cell. By molecular standards, they are giant — roughly 100 times more massive than even very large conventional polymers and the largest known proteins in biology. Yet despite their enormous size, scientists can precisely control their lengths and organize them into chemically distinct segments with defined dimensions carrying opposite electrical charges.

That precision also produced an unexpected biological effect. In cell culture studies, the structures enhanced the growth and organization of neurons and increased synapse formation.

The findings introduce a new way to use self-assembly to build highly controlled supramolecular materials with precise physical arrangement of chemical structures. Ultimately, the work could inform materials’ design for a range of applications, including biomaterials for regenerative medicine.

“These structures represent a breakthrough in materials design, and we already discovered one useful application: their superbioactivity toward neurons, which will bring new opportunities in regenerative medicine,” said Northwestern’s Samuel I. Stupp, who led the study. “Typically, when we design a regenerative material, we add biological signals, designed to activate cell receptors. Here, the material itself becomes highly bioactive simply through the precise organization of electrical charges and its dynamic behavior. That introduces a very different way of thinking about how materials can communicate with cells.”

Bringing precision to self-assembly

The new study builds on Stupp’s long history working with supramolecular materials, dynamic structures that form when molecules spontaneously organize through weak, reversible interactions. While conventional polymers, like plastics, hold together using strong, permanent chemical bonds, supramolecular polymers are dynamic, allowing their components to move and reconfigure, which generates life-like properties in synthetic systems. Supramolecular polymers exist in all living organisms, where they perform essential functions and hold promise for future soft matter technologies.

That flexible, dynamic nature can be useful for interacting with living cells in biomedical applications, but it also makes the structures difficult to control. Scientists can design the individual molecular building blocks but then struggle to control how long the resulting structures grow and how they organize their chemical components.

Stupp’s team overcame that limitation by using novel designs of molecules and carefully controlling their assembly process. With this strategy, the researchers could dictate the threads’ lengths and keep them uniform. Once the growth stopped, the structures maintained their precise lengths rather than fusing together or rearranging.

“We let these filaments sit in a test tube for months, and they surprisingly didn’t change,” Stupp said. “They remained the same length and retained their segmented structures.”

Grow, stop, grow again

Stupp and his team traced that unusual ability to a previously unknown process his laboratory calls “self-capping supramolecular polymerization.” As a thread grows, flexible portions at the ends remain highly dynamic and, at proper temperatures, allow new molecules to join in a synchronized manner. Once the available building blocks are consumed, those flexible portions fold over and protect the ends. This effectively shuts down further growth by preventing threads from fusing together or undergoing “Ostwald Ripening,” a process in which smaller structures disappear while larger ones grow.

“To make them longer and longer, we just kept adding new molecules,” Stupp said. “The new molecules stick with strong noncovalent bonds to the growing thread and elongate the assembly, which led to a breakthrough in terms of size control.

“The collective molar mass of their noncovalently bonded monomers reaches billions of daltons, and thus they are ‘giga-assemblies.’ These are the largest supramolecular polymers ever made with this degree of precision. Once the monomer is fully consumed, they finish growing and their dimensions remain stable.”

Approximately the size of one proton or neutron, a dalton is a tiny unit that scientists use to measure the mass of molecules and atoms. More than a billion daltons is enormous by molecular standards. For comparison, the largest common polymers of plastics have molar masses in the range of 10 million daltons, and the largest proteins in nature reach approximately 3 to 5 million daltons.

Researchers can later restart growth by supplying more building blocks. By adding different types of molecules in sequence, the researchers can create precisely defined segments with different chemical properties. In the study, Stupp and his team created threads with positively and negatively charged segments. Because opposites attract, the researchers expected the threads to clump together. Instead, they remained stable in water because the “capping” mechanism prevented the assemblies from fusing together.

Small anchor, big difference

To explore how the pattern of electrical charges might interact with living cells, Stupp and his team turned to neurons, which carry a strong negative charge. The team exposed several threads with different designs to cellular cultures containing cortical neurons.

“The first thing we did was make a structure with only positively charged segments, and it basically killed the neurons,” Stupp said. “Then, we tried a structure with only negatively charged segments, and nothing happened because the neurons repelled the materials.”

When Stupp’s team used supramolecular assemblies with randomly mixed positive and negative charges, the neurons didn’t respond in a notable way. But then the team designed a thread with the breakthrough polymerization strategy comprising a small positively charged segment sandwiched between two negatively charged segments of different lengths. Something interesting and unexpected occurred.

The small positive region anchored the thread to the negatively charged surface of a neuron. Repelled by the neuron’s negative charge, the thread’s negative segments remained highly mobile and in constant motion since they could not escape.

“When we used only positively charged segments, they suffocated the neuron and killed it,” Stupp said. “But with only one positively charged segment surrounded by negative ones, it couldn’t do that anymore. And the negative segments, which ordinarily couldn’t get near the neuron, were forced to stay near it because they were anchored.”

Neurons grow and spring to life

Neurons treated with the precisely segmented threads grew longer neurites — the long projections neurons use to communicate with other cells. The treated neurons also formed more elaborate branches, gaining the ability to connect with neighboring cells.

“The neurons matured to an amazing level,” Stupp said. “They grew long and highly branched neurites, which is how they can start making synaptic connections in the brain and in the spinal cord.”

After just seven days, treated neurons were much more active than untreated neurons. When stimulated, treated neurons produced a calcium response — a sign of neuronal activity. Their overall calcium activity was four times greater. And, after 14 days, treated neurons formed more synapses than untreated neurons.

Stupp and his team saw another striking effect in cultures of human neural progenitor cells, which are immature cells that can develop into neurons. Although these cultures typically grow in flat layers, the cells instead organized into raised, three-dimensional clusters connected by axons. This organization resembled aspects of the brain’s white and gray matter, where somas are separated from axons.

No signals required

Perhaps most surprisingly, the molecular threads produced these effects without carrying biological signals specifically designed to encourage neurons to grow. Stupp posits the negatively charged segments recruit specific proteins already present around the cells that help neurons survive, grow and communicate. These proteins, called neurotrophic factors, have positively charged domains. Stupp’s team also found laminin — an important protein that holds tissues together and has positively charged domains — accumulating along the supramolecular threads near neurons.

Stupp hypothesizes that the interplay between attraction and repulsion causes the threads’ negative segments to rapidly move toward and away from the neurons in a “tapping” motion. Proteins recruited by the negative segments may be drawn toward receptors on the neurons, while electrostatic forces push the segments away. This repeated tapping, occurring on millisecond timescales, could help enhance biological signaling. A live cell movie included in the Science paper has captured this phenomenon.

Beyond neurons, Stupp says the new polymerization strategy could eventually enable scientists to build long molecular structures containing many carefully positioned chemical segments. That could open opportunities to create new materials and devices with features that extend over very long distances.

“We discovered something we had never seen before in chemistry or in biology,” Stupp said. “This work gives us a new level of control over supramolecular materials. In the future, we can imagine creating many segments with different properties. The unknown functions of these structures are likely to be as surprising as the superbioactivity observed on neurons. We’re just beginning to imagine the possibilities.”

The post Supramolecular polymers can dramatically enhance brain cell activity appeared first on Mapping Ignorance.

Yu–Shiba–Rusinov states

Magnetism is an old enemy of this order. Magnetic impurities were known to suppress superconductivity, and in the 1960s Yu, Shiba and Rusinov showed theoretically that a magnetic atom disturbs the pairs around it and traps a localized state with an energy inside the gap. These Yu–Shiba–Rusinov (YSR) states were first seen atom by atom in 1997, using a scanning tunneling microscope (STM): a needle-sharp metal tip that measures the tiny current of electrons tunneling across a vacuum gap to a surface.

A different story concerns two superconductors joined by a very weak link. In 1962, Brian Josephson predicted that Cooper pairs can tunnel through such a link with no voltage at all, an effect confirmed in 1963. When a voltage is applied, another process takes over: multiple Andreev reflection. An electron whose energy lies inside the gap cannot enter a superconductor alone. Instead it is sent back as a hole, the absence of an electron, while a Cooper pair slips in, as Andreev described in 1964. Between two superconductors under a voltage V, a particle can bounce back and forth, gaining energy eV with each crossing, until it can finally escape above the gap. When n crossings are needed, the conductance shows a peak at a voltage equal to the sum of the two gaps divided by n. This pattern, explained in the early 1980s, is called subharmonic gap structure. For lead on lead the sum is 2.7 millivolts, so peaks are expected near 2.7, 1.35 and 0.9 millivolts.

Impurity states and Andreev reflections have usually been treated as separate subjects. A recent experiment ¹ places them in the same device. A nickelocene molecule, a nickel atom sandwiched between two five-carbon rings, sits on a lead surface above a single iron atom trapped between the molecule and the lead. On its own, the molecule carries a spin that appears as characteristic steps in the measured spectra. Above the iron, those steps vanish and YSR states appear instead. Density-functional calculations explain why: the magnetic moments of molecule and iron point in opposite directions and almost cancel, leaving a leftover spin of one half on the iron. That spin couples to the lead and produces a single pair of YSR states.

The second superconductor is a lead-coated STM tip, lowered toward the molecule at 2.4 kelvin. Far away, electrons tunnel through a wide vacuum barrier and the tip merely reads out the YSR states, which appear as lopsided peaks near the edge of the gap. As the tip approaches, the junction becomes more transparent, meaning electrons cross it more easily, and the setup becomes a weak link between two superconductors. At moderate conductance, new peaks appear at half the sum of the gaps, the signature of second-order Andreev reflection. At the highest conductances, a sharp peak develops at zero voltage: the Josephson contribution. Thermal fluctuations of the superconducting phase broaden this peak, as is typical of STM junctions.

These stages do not simply stack. As the tip approaches, the YSR peak moves to lower energy and mingles with the Andreev peaks. The interpretation is that the bound state becomes a stopping point along the Andreev path. Peaks then appear at the tip’s gap plus the bound-state energy, divided by n, rather than at the ordinary positions. This is called YSR-assisted multiple Andreev reflection.

Odd and even orders

A simple argument explains why odd and even orders differ. Each reflection carries the particle to the opposite electrode, and only the lead surface contains the magnetic atom; the tip has none. After an odd number of crossings the particle ends on the surface, where it can land in the bound state. The peak shifts and acquires the lopsided shape typical of tunneling into a single level. After an even number it ends in the tip, among ordinary states, and those peaks stay closer to their usual positions, although the second-order peak still turns lopsided at higher conductance as a YSR-assisted component splits off. Simulations reproduce this pattern. In the measurements, thermal broadening blurs finer details such as the third-order peak, and small features near zero voltage seem to reflect a mixture of both processes.

Reproducing the evolution of the spectra required two changes together: higher transparency and a stronger exchange coupling between the iron spin and the lead. Raising transparency alone gives a growth of the key peak that the measurements do not show. A likely reason is that the approaching tip slightly squeezes the molecule and strengthens the coupling, though this is an inference. The model treats the spin as classical and its two parameters as effective ones, so it is a phenomenological decomposition rather than a unique microscopic extraction. It also cannot capture the zero-voltage peak.

Even so, the result shows that YSR states and Andreev reflections are two limits of one problem, tuned continuously within a single molecular junction. Under suitable conditions, such spin-active links might also host a Josephson diode, in which supercurrent flows more easily in one direction, or a π-junction, in which the two superconductors prefer opposite phases. These remain possibilities, not demonstrated effects.

Author: César Tomé López is a science writer and the editor of Mapping Ignorance

Disclaimer: Parts of this article may have been copied verbatim or almost verbatim from the referenced research paper/s.

The post How a magnetic molecular junction shapes the current between two superconductors appeared first on Mapping Ignorance.

 

Astronomers have catalogued billions of stars and hundreds of billions of galaxies. We have found thousands of planets orbiting stars. We can also map where these galaxies sit and trace how the universe changes over time.

The universe also contains objects that are too faint, too far away, or too crowded to study easily. Simply taking a picture of a galaxy tells us only parts of its story.

When we look at breathtaking pictures of the universe, we rarely see the hidden limits of the machine that built them. Even the best telescopes have mechanical blind spots that can quietly miss key galaxies. If we ignore these missed targets, they can affect our understanding of how galaxies and other cosmic objects evolve over time.

Our new paper, published in the Astronomical Journal, presents a computer tool to show how clever instrument design can solve these hardware limitations. It models how tiny telescope robots work so we can catch missed targets to map the universe accurately.

Thousands of tiny robots

Modern astronomy is not just about individual stars and galaxies.

Instead, scientists design massive surveys to scan large areas of the sky multiple times. The bigger these surveys are, the easier it is to learn how stars form, how galaxies grow, and how the universe changes over time.

Modern cosmic survey telescopes use thousands of tiny robots to position optical fibres to collect light from distant stars and galaxies. Future telescopes plan to use over 20,000 or more of these at once.

Each robot’s optical fibre directs the light from a distant star or galaxy to a tool called a multi-object spectrograph. This splits the light to show astronomers what the galaxy is made of and how far away it is.

The tiny robots can do a great job, but they have physical limits. They can bump into each other. They can’t get too close to each other without crashing. In crowded parts of the sky where galaxies sit close together in dense clusters, the robots can run out of room.

As a result, the system skips some galaxies to avoid crashes. These missing galaxies can create blind spots. If we use these biased data our maps can give us misleading answers about the universe.

Finding the blind spots

To keep the robots safe, engineers set strict rules such as the space between robot bases (called pitch), how far each robot can move around its base (called patrol radius), and the safety gap two robots must keep so that they do not touch (called exclusion radius).

To understand how these constraints introduce blind spots, our new study introduces a proof of concept computer tool which acts as a testing ground before a telescope turns on.

Our software can test many combinations of robot systems and astronomical targets. It acts like a flight simulator for telescopes. It mimics the process of moving a robot to position its optical fibre to many targets over repeated visits to the sky.

This tracks how well the system assigns fibres to targets, called allocation efficiency. It also tracks how many total galaxies get observed over time, called survey completeness.

In our study, we ran tests using specifications from real telescopes. Our results showed three key rules.

First, setting the smallest pitch, or using the highest fibre density, gives the best completeness and efficiency.

Second, when the pitch is fixed, increasing how far a robot can reach is the next best way to improve fibre assignment.

Third, making the safety gap smaller also helps stop missed targets, but it has a smaller effect than changing the reach of the robot.

Our tool bridges the gap between science goals and machine limits. By testing these rules early, astronomers can identify and address technology limits in designing new instruments. Engineers can use our code to test new ideas for future instruments before spending millions of dollars to build them.

There is still much of the universe that we haven’t mapped yet. Future maps of the universe will depend on more than just giant telescope mirrors. They will also need smart computer tools that improve how much telescopes can actually see.

Sometimes, mapping the whole universe starts with getting thousands of tiny robots to move in exactly the right way.

 

This article is republished from The Conversation under a Creative Commons license. Original article.

The post Even the best telescopes have blind spots appeared first on Mapping Ignorance.

The study ¹ demonstrates that waves inside an irregularly shaped cavity made from hyperbolic materials (which are named after the hyperbola mathematical curve because of the unique shape these materials force light waves to take) can organize into stable, repeating paths rather than scattering chaotically. The researchers call these structures “hyperbolic wave attractors,” and their findings could eventually help scientists and engineers design new ways to control light, radio waves, and sound in complex environments.

“This work shows how geometry and the properties of a material can work together to produce wave behavior that is both surprising and useful,” said Andrea Alù, the study’s principal investigator, director of the Photonics Initiative at the CUNY ASRC, and Distinguished Professor of Physics at the CUNY Graduate Center. “By understanding how waves organize themselves in these hyperbolic media, we can begin to explore new approaches to controlling energy, information, and communication signals in complex environments.”

When waves take an unusual turn

In everyday materials, waves generally reflect from a surface in a familiar way: The angle at which a wave arrives matches the angle at which it leaves. In an irregularly shaped room, repeated reflections can send waves in many directions, creating complex chaotic patterns. This is the basis of a classic physics problem known as a dynamical billiard, in which the motion of a ball — or, in the wave version, light — inside a curved or irregular container quickly becomes unpredictable.

Hyperbolic materials can drastically change this picture. These materials have unusual properties that force waves to travel along narrow, highly defined directions instead of spreading freely in all directions. As a result, when a wave encounters a tilted wall, its outgoing direction can differ from what would be expected in an ordinary material.

Alù’s team wanted to understand what would happen when these unusual reflection rules were combined with an irregularly shaped cavity.

“Normally, we expect a complicated cavity to produce complicated, chaotic wave patterns,” said Simon Yves, a postdoctoral researcher in Alù’s lab and a first author of the study. “Here, the opposite happens. The unusual propagation and reflection of waves create a strong geometric organization, producing well-defined paths that can persist across a broad range of wavelengths.”

From chaos to wave attractors

The researchers found that the waves inside their cavity can progressively organize into closed trajectories. These paths are stable and scale-invariant, meaning their underlying geometric structure persists across different scales.

The effect arises from a breaking of mirror symmetry in the wave-reflection process. As waves bounce around the cavity, their wavelengths progressively shrink, helping create the conditions for the formation of the attractors.

The resulting wave patterns also have a property called handedness. This describes whether the defined wave’s trajectory inside the cavity rotates clockwise or counterclockwise. The attractors’ handedness and stability are connected to the unusual geometry of wave propagation in the hyperbolic material.

The team demonstrated these effects by engineering vibrations in a mechanical metamaterial, a human-made structure designed to control how waves move through it.

“The most exciting aspect of these results is that they connect a simple geometric idea with a rich set of wave phenomena,” said Enrico Renzi, a doctoral student in Alù’s lab and a first author of the study. “We can observe how the waves become organized, and we can connect that organization to properties such as stability and handedness. This gives us a framework for designing wave behavior rather than simply observing it.”

A bridge from ocean waves to nanophotonics

The researchers’ findings have intriguing parallels with internal wave attractors studied in oceanography and fluid dynamics. In oceans and other stratified fluids, waves traveling through water with density gradients can reflect unusually from underwater slopes. These reflections can focus energy into closed paths and contribute to wave turbulence.

The new study translates a related phenomenon into a tabletop, solid-state metamaterial platform, showing how concepts from geophysical fluid dynamics can be explored in engineered materials and in simple linear settings.

The framework that the scientists introduced can also extend to hyperbolic phonon polaritons — hybrid light-matter excitations that can arise in natural two-dimensional anisotropic crystals such as hexagonal boron nitride and molybdenum trioxide, which support similarly odd propagation and reflection properties. These materials can support light confined to tiny regions, offering opportunities for low-loss nanophotonic circuits, enhanced interactions between light and matter, and control of infrared light.

In the future, hyperbolic wave attractors could inform the design of compact optical chips, devices that separate optical signals, particle-trapping systems, analog wave-computing platforms, and highly sensitive biological sensors.

“This research opens a path toward engineering stable and robust wave patterns in systems where waves would normally be expected to rapidly become chaotic,” Alù said. “Our goal is to understand how these effects can be harnessed to create new functionalities for photonics, phononics, and wave-based technologies.”

The post Waves find order in the chaos of an oddly shaped cavity appeared first on Mapping Ignorance.

26 April 2026 will always be remembered as the day an apparently impossible barrier was finally broken. At the London Marathon, Kenya’s Sabastian Sawe and Ethiopia’s Yomif Kejelcha became the first men to officially run the 42.195 km distance in under two hours, finishing in 1:59:30 and 1:59:41 respectively. Reaching this landmark was not the result of a single innovation but of decades of progress in training, nutrition and sports science. One piece of technology, however, has increasingly come into focus: the shoes beneath the athletes’ feet. Modern racing shoes combine lightweight foams with rigid carbon-fibre plates and carefully engineered geometries to improve running efficiency and energy return. Similar advances can be seen in many other sports. In cycling, for example, bicycles and sportswear are becoming lighter and more aerodynamic, with components engineered to combine low weight, stiffness and reliability. In mountaineering, waterproof and breathable membranes allow climbers to cope with extreme weather conditions by controlling the transport of moisture through their clothing. In all these cases, performance increasingly depends not only on what a material is made of but also on how it is structured. With environmental challenges growing, this raises another question: could the next generation of high-performance materials also be more sustainable?

Cellulose is found in everything from wood and cotton to agricultural residues and other forms of biomass, and it gives plants the mechanical strength they need to grow. But cellulose is not interesting simply because it is abundant. As noted above, the way a material is structured can be just as important as its chemical composition. At the nanoscale, this becomes particularly striking: the same cellulose molecules can be organised into structures with very different dimensions, morphologies and properties. This ability to modify a material’s properties by controlling its structure at the nanoscale is the basis of nanotechnology, and cellulose is not the only material in which it can be exploited. Quantum dots, for example, are semiconductor nanostructures whose optical properties change with their size. They are now used in technologies such as QLED displays. In a similar way, controlling cellulose at the nanoscale can transform this common biopolymer into a material with entirely new capabilities. This is the idea behind nanocellulose. The term does not describe a single material but a family of nanoscale forms of cellulose (Figure 1), mainly cellulose nanocrystals (CNCs), cellulose nanofibres (CNFs) and bacterial nanocellulose (BNC). Although they share the same basic chemical building blocks, their different structures and origins give them distinct properties and potential applications.

But how can cellulose be brought down to the nanoscale? For CNCs and CNFs, the starting point is cellulose in a much larger form, such as wood pulp, cotton or other plant fibres. Through physical, chemical and/or biological treatments, these structures can be progressively disintegrated into their nanoscale building blocks. The production of CNCs selectively removes the less ordered regions of cellulose, leaving behind short, rigid and highly crystalline nanocrystals. CNFs, in contrast, are obtained by breaking the fibres down into long, flexible nanofibrils that can form interconnected networks. These structural differences give each type its own strengths: CNCs can stabilise emulsions, reinforce polymers or help disperse other nanomaterials, while CNFs can form films, hydrogels and lightweight aerogels. The surface of cellulose can also be chemically modified in many different ways, allowing its properties and its interactions with other materials to be tailored for specific applications. BNC takes a different route. Rather than breaking down plant cellulose, bacteria build cellulose directly from sugars such as glucose or sucrose, as a network of nanoscale fibrils. Komagataeibacter xylinus, for example, is one of the species found in the bacterial community used to make kombucha. Growing at the interface between the culture medium and the air, these bacteria produce a highly pure, crystalline and porous cellulose network that can be used in membranes, biomedical scaffolds and other three-dimensional materials.

The potential of nanocellulose stems from the unusual combination of properties that can emerge from these different nanoscale structures, and researchers are exploring an increasingly broad range of uses. CNCs, for example, can stabilise interfaces between otherwise incompatible materials, making them useful in emulsions, while their reinforcing ability is being investigated for making lighter and stronger polymer composites. Their surface chemistry also makes them attractive as dispersing agents for other nanomaterials, opening up possibilities for water-based functional inks and coatings. CNFs, with their long and interconnected fibrils, can instead form continuous networks that can retain large amounts of water or be dried into lightweight, porous structures. This has led to research into films, hydrogels and aerogels, as well as three-dimensional materials for insulation, separation and other applications. BNC offers another distinctive architecture: its nanofibrillar networks can be produced directly as membranes and are being investigated for biomedical materials and other applications where controlled porosity and water management are important. Beyond these more established research directions, nanocellulose is also being explored as a platform for sensors, conductive materials, energy devices and electronic textiles, often by combining it with other functional materials. The important point is that nanocellulose is not simply cellulose in a smaller form: nanoscale organisation gives this abundant biopolymer new functions that are difficult to achieve with conventional materials.

Yet the potential of a material cannot be measured by functionality alone. Its environmental performance must also be considered: how much energy and raw material are needed to produce it, what happens during use and whether it can be reused, recycled or safely disposed of afterwards. Major sports brands are already exploring new materials and manufacturing approaches aimed at combining high performance with a lower environmental impact, driven by rising consumer awareness and tighter environmental regulation. Replacing a fossil-based material with one derived from biomass does not automatically make a product sustainable if its production is energy-intensive, its lifetime is short or its end of life is problematic. The real challenge is to design materials whose functional advantages are accompanied by a lower environmental burden across their entire life cycle. This is where nanocellulose becomes particularly interesting: it is a renewable alternative to conventional materials and a platform for new functionalities, with the potential to enable more biodegradable and environmentally benign materials.

Nanocellulose will not replace every material, nor should it. Its value lies in offering another set of possibilities for designing materials in which performance and environmental considerations can be addressed together. The journey from a familiar polymer found in wood and cotton to nanoscale crystals, fibres, networks and three-dimensional structures illustrates how much can change when we learn to control matter at ever smaller length scales. The question for the future is therefore not simply whether we can make materials from renewable resources but whether such materials can be genuinely better – and better for the world around us.

References

1. Víctor Calvo, Carlos Martínez-Barón, Laura Fuentes, Wolfgang K. Maser, Ana Benito and José Miguel González-Domínguez (2024). Nanocellulose: the ultimate green aqueous dispersant for nanomaterials. Polymers. doi: 10.3390/polym16121664.
2. Qinghua Ji, Cunshan Zhou, Zhenqi Li, Isaac Duah Boateng and Xianming Liu (2023). Is nanocellulose a good substitute for non-renewable raw materials? A comprehensive review of the state of the art, preparations, and industrial applications. Industrial Crops and Products. doi: 10.1016/j.indcrop.2023.117093.

The post Nanocellulose: from the most abundant biopolymer to next-generation sustainable materials appeared first on Mapping Ignorance.

One of the most powerful ways to study the Universe is to map where galaxies are located. Galaxies are not scattered randomly through space; they gather into groups, clusters, filaments and other structures that trace the underlying distribution of dark matter. The pattern of this cosmic web carries information about the basic ingredients and history of the Universe.

Pushing into a small-scale regime

A new analysis ¹ pushes this idea into an unusually small-scale regime. It focuses on emission-line galaxies, a population of relatively young, actively star-forming galaxies. These objects are especially useful for surveys such as the Dark Energy Spectroscopic Instrument, or DESI, because their strong spectral emission lines make it comparatively easy to measure how far away they are, and therefore how far back in cosmic time they are being seen. The data used here come from DESI’s first public data release, covering galaxies seen when the Universe was between roughly 30 and 50 percent of its current age, some seven to ten billion years ago.

The difficulty is that galaxies do not trace matter in a simple, one-to-one way. Most of the matter in the Universe is dark matter, and galaxies form and live inside concentrations of it called halos. How a galaxy relates to its halo depends on messy astrophysical processes, such as how efficiently it forms stars, and what happens when it falls into a larger halo, loses gas, and gradually shuts down its star formation. These complications become especially important for galaxies separated by only a few million light-years or less, distances comparable to the size of a single halo.

For this reason, cosmological analyses have traditionally focused on larger separations, where the growth of cosmic structure can be described with simpler mathematics. But this throws away information. The smallest scales contain far more detail about how matter is arranged inside halos, and in principle they can provide much tighter constraints on cosmological parameters. The challenge lies in separating that detail from the astrophysics of galaxy formation itself.

A model that links galaxies to the dark-matter structures

The new analysis tackles this problem with a model (SHAMe-SF, Subhalo Abundance Matching extended – Star Forming) that links galaxies to the dark-matter structures found in computer simulations. Rather than assuming every halo hosts galaxies in exactly the same way, the model lets several properties vary. It can, for instance, distinguish central galaxies from smaller satellite galaxies orbiting within the same halo, and it can capture the gradual suppression, or quenching, of star formation in some environments. The calculations rely on large cosmological simulations, sped up with a machine-learning shortcut, so millions of parameter combinations can be explored efficiently.

Before turning to real observations, the method was tested on artificial galaxy catalogues built from a large hydrodynamical simulation, in which the true underlying cosmology is known in advance. Two different mock populations of emission-line galaxies were constructed from this simulation. In both cases, the method recovered the correct cosmological parameters without significant bias, using clustering information down to separations of only about one and a half million light-years, deep inside individual dark-matter halos.

Smallest scales are not just extra data points

These tests revealed something particularly interesting: the smallest scales are not just extra data points. They can be essential for getting the cosmology right. The clustering signal inside individual halos carries information about satellite galaxies, and if those scales are discarded, the properties describing satellites become tangled up with σ8, the parameter that measures how strongly matter clusters throughout the Universe. In the simulations, excluding separations below roughly three to four million light-years produced looser, and in one case biased, estimates of that parameter.

The method was then applied to real data: the lower-redshift half of the DESI emission-line-galaxy sample. Three complementary measures of galaxy clustering were combined, capturing both how galaxies are separated across the sky and how their apparent positions are stretched or compressed by their own motions along the line of sight. The smallest separations included were again about one and a half million light-years.

Striking agreement

The result is σ8 = 0.81, accurate to roughly six percent. This number describes how strongly matter is clumped together: higher values mean matter is bunched more tightly into structures, lower values a smoother distribution. Alongside it, the analysis measures Ωm h² = 0.146, also accurate to about six percent. This quantity combines the fraction of the Universe made of matter with its expansion rate, giving an absolute measure of how much matter it contains, independent of that rate. Both values agree with measurements of the cosmic microwave background made by the Planck satellite, and with other DESI analyses. For comparison, an independent DESI study that combined the clustering of several galaxy types across the survey’s full volume, with added information about the density of ordinary matter and the shape of the primordial density fluctuations, found σ8 = 0.842, with an uncertainty of about four percent.

What is striking is not simply this agreement. The measurement described here comes from only one percent of the DESI survey’s eventual volume, and it relies on scales usually considered too complicated to model reliably. Yet its precision rivals that of analyses built on the full, much larger dataset.

There is an important qualification. The study does not prove that the completed DESI survey will automatically deliver ten times better precision, even though a simple statistical extrapolation suggests such a gain might be possible. Systematic effects could well become the limiting factor. The galaxy-halo model needs testing against a wider range of simulations, effects such as which galaxies actually get their light measured need to be better understood at small scales, and the computational shortcuts need to become even more precise.

The broader lesson is that the nonlinear Universe, the tangled, complicated regime close to individual galaxies, is not merely a mess to be avoided. If its astrophysical complexity can be modelled reliably, the intricate distribution of galaxies inside dark-matter halos becomes a genuine source of cosmological information. What was once treated mainly as a complication may turn out to be one of the most valuable parts of the cosmic map.

Author: César Tomé López is a science writer and the editor of Mapping Ignorance

Disclaimer: Parts of this article may have been copied verbatim or almost verbatim from the referenced research paper/s.

The post Reading the cosmic web in its smallest scales appeared first on Mapping Ignorance.

Right now, your brain is recognising the shapes of these letters, analysing their meaning, and holding a stream of thought together – all on about 20 watts of power. That’s roughly what it takes to run a lightbulb.

Yet reproducing even a small fraction of the brain’s capabilities using today’s AI systems can require vast computing infrastructure and energy.

Neuromorphic computing – also known as “brain-inspired computing” – is an attempt to close that gap. Not by making today’s computers slightly more efficient, but by fundamentally rethinking how they should work.

The human brain holds roughly 86 billion neurons, wired to each other through junctions called synapses. A neuron receives signals from its neighbours; when their combined effect crosses a threshold, it will fire a brief electrical pulse known as a spike.

Two things make this process efficient. First, energy is spent only where it’s needed, not everywhere constantly. Second, memory and processing are intertwined in the strength of each synapse’s response. The brain doesn’t have to pause and go find information, saving both time and energy.

In contrast, in an ordinary computer chip, the processor and memory are physically separate, so every calculation entails transferring data between them. This can account for a substantial share of energy use – known as the Von Neumann bottleneck, a term first used by US computer scientist John Backus in 1978 to describe this computing inefficiency.

Neuromorphic computing tackles this inefficiency in several ways: bringing memory and processing closer together, using “sparse representations” (data models where most values are zero), and performing computation only when events occur.

This is known as event-driven computing – and we are already seeing some exciting applications.

Sensors modelled on human retina

Event cameras are sensors modelled on the human retina. Their pixels respond individually and only when they detect a change in the scene – rather than capturing a full frame dozens of times a second, like a smartphone or video camera.

The result is a neuromorphic sensor that uses a fraction of the power, handles fast motion without blur, and works equally well in bright sunlight and near-total darkness.

In autonomous vehicles, this combination of low latency and reliable performance in glare or darkness could be the difference between detecting a pedestrian in time, or not. And in space, where power is scarce and lighting extremes are common, event cameras have already been deployed for object and debris tracking.

It might be tempting to ask if neuromorphic hardware could do for computing what Nvidia’s graphics processing units (GPUs) have done for AI. But these are fundamentally different kinds of chip.

Nvidia’s GPUs are exceptionally good at the dense, repetitive maths behind deep learning, which has made them the default choice for almost any AI task. In contrast, neuromorphic computing’s event-driven approach looks likely to spawn a wide range of specialised chips, rather than a single winner-takes-all design that everyone adopts at once.

Australia’s BrainChip already sells a commercial neuromorphic processor for powering cameras and sensors that need to run at very low power. Expect neuromorphic hardware to sit alongside conventional processors and GPUs rather than replacing them, taking on the jobs where its efficiency edge is decisive.

Sensitive information

But there is another advantage to neuromorphic computer technology beyond energy saving.

Because neuromorphic chips can process data right where it’s generated, sensitive information from wearables, smart cameras, and medical sensors can stay on the device instead of travelling to the cloud, potentially reducing privacy and cybersecurity risks associated with data transmission.

Devices can also work offline, without needing to send data to a distant server and wait for a reply. Cutting out that round trip matters wherever split-second decisions are required, such as in autonomous vehicles, drones, and robots operating beyond reliable signal.

This won’t make data centres redundant. But fewer tasks may need one in future, and those that remain could become dramatically more efficient. IBM has already demonstrated energy savings of up to 10,000-fold on event-driven tasks, compared with conventional digital architectures, with its TrueNorth chip.

Rethinking how computing works

Computing has been through a transition like this before. The first computers filled entire rooms and drew as much power as a small factory. Today, however, a chip smaller than a fingernail in the phone in your pocket is far more capable.

Neuromorphic computing raises the possibility of a similar shift – driven less by cramming in more transistors (the tiny switches that make up a chip’s circuitry) than by rethinking how computing should work in the first place.

But turning this promise into everyday infrastructure will take sustained effort. At NeuroWare, the UK’s leading multi-university innovation hub for neuromorphic computing, I have led the development of a UK roadmap for this technology up to 2050, working with colleagues across academia, industry and government.

With the global market for neuromorphic technology forecast to nearly quadruple to US$20 billion (£14.8 billion) by 2030, the UK needs open-access facilities to prototype new chips, coordinated investment, common standards – and a workforce trained to bridge neuroscience, electronics and computer science.

The foundations are already here. What’s missing is the infrastructure to take neuromorphic computing from supercomputers such as the University of Manchester’s brain simulator, SpiNNaker, to the billions of ordinary devices that could one day use brain-inspired hardware for themselves.

Your brain solved this problem millions of years ago. It’s taken science a long time to appreciate just how well it works.

 

This article is republished from The Conversation under a Creative Commons license.  Original article.

The post My brain runs on 20 watts. Take that, AI ! appeared first on Mapping Ignorance.

Somewhere between the microscopic realm of elementary particles and the macroscopic world of human beings, something strange happens: The rules of quantum physics, which work so exquisitely for tiny atoms, seem to lose their grip as objects grow larger. Pondering where and how this shift from small-scale quantum fuzziness to everyday sharp certainty happens gives rise to thought-experiment oddities like Schrödinger’s famous dead-and-alive cat. The process by which quantum phenomena like superposition—the paradoxical affliction of Schrödinger’s cat—fade into the classical reality we experience is known as decoherence. Now, a new FQxI-funded experiment has narrowed the field of possible explanations for decoherence, in particular ruling out a prominent theory linking gravity to the process. ¹

“One of the deepest questions in modern physics is why the strange quantum behavior that governs atoms and elementary particles seems to disappear in the macroscopic world we experience every day,” says FQxI member Catalina Curceanu, director of research and spokesperson for the VIP Collaboration at the National Laboratory of Frascati of the National Institute for Nuclear Physics (INFN-LNF) in Italy.

Conducted at the INFN Gran Sasso National Laboratory (INFN-LNGS), the world’s largest underground laboratory for fundamental physics research, the experiment tested one model in which decoherence is caused by gravity. Einstein’s general theory of relativity states that gravity manifests due to the warping of spacetime’s fabric around massive objects. In the 1960s, the Hungarian theoretical physicist Frigyes Károlyházy posited that spacetime is constantly rippling with tiny fluctuations that gradually erode quantum superpositions, preventing macroscopic objects from existing in the kind of quantum combinations imagined in Schrödinger’s famous cat paradox. His model continues to intrigue physicists and was recently revived, refined and reformulated by FQxI’s Angelo Bassi and colleagues.

Telltale trails

The fluctuations predicted by Károlyházy can’t be observed directly but, if they exist, they should cause charged particles to jiggle and accelerate randomly, giving telltale trails of electromagnetic radiation. This radiation would be so faint that it could easily be lost in electromagnetic background noise from sources like cosmic rays. That makes the Gran Sasso National Laboratory, which is tucked beneath 1.4 kilometers of radiation-dampening rock, an ideal place to conduct the search. “The natural shielding provided by the rock creates one of the quietest environments on Earth for detecting extremely rare physical phenomena,” says Curceanu.

The researchers used a detector made up of a coffee-mug-sized piece of high-purity germanium crystal, surrounded by layers of copper and lead shielding. They collected data for a total of 62 days. Then, they subtracted the expected background radiation from their measurements and looked for a signature that matched that predicted by the model.

The result: No signal.

This doesn’t entirely rule out the possibility that gravity plays a role in quantum decoherence. But it does provide important information about where to look for a possible gravitational link. “This absence of a signal is itself a major scientific result,” says Curceanu. “By ruling out one of the oldest and most natural gravity-induced decoherence models, this work narrows the search for the theory describing the interplay between gravity and quantum mechanics, bringing us one step closer to understanding one of the deepest mysteries in fundamental physics.”

Exiting the realm of speculation

Károlyházy’s model rests on the notion that there is a fundamental limit to the precision with which we can locate objects and measure length. In the years since he proposed the model, this feature has emerged as a common thread in many contemporary theories seeking to unite quantum physics and gravity, including string theory and loop quantum gravity. “Every quantum gravity approach ends up with predicting the existence of a minimal length connected to the uncertainty in the measurement of spacetime,” says Kristian Piscicchia, a quantum physicist at the Enrico Fermi Research Center/INFN/VIP, in Italy, and the experimental lead on the new study.

Although many assume that quantum gravity cannot be probed by current technologies, the new research joins a growing body of work demonstrating that some ideas that involve both gravity and quantum theory are testable today. “Precision experiments are now reaching a level of sensitivity where they can test ideas that, until recently, belonged almost exclusively to the realm of theoretical speculation,” says Curceanu. “As sensitivity improves, the boundary between theory and measurement continues to move, opening new possibilities for discovering the fundamental principles that govern our universe.”

The post What kills Schrödinger’s cat? appeared first on Mapping Ignorance.

Inactivation of tumor suppressor genes (TSGs) confers a cellular fitness advantage to human cancer cells, including in acute myeloid leukemia (AML), where it disrupts myeloid differentiation and promotes leukemogenesis. Loss-of-function mutations in TSGs are frequently observed in patients with hematologic malignancies ¹, driving disease progression by blocking differentiation and enhancing self-renewal. Therefore, restoration of tumor suppressor activity represents a promising therapeutic strategy ². Indeed, previous studies indicate that the therapeutic potential of this approach lies in the reversibility of TSG silencing ³. In the present work, Arnuk A and coworkers  identified the transcriptional repressor ZBTB7A as a TSG that is downregulated in AML, especially within leukemia stem cells (LSCs). Its loss impairs differentiation and acceleraties disease progression through aberrant activation of inflammatory pathways ⁴. Importantly, the authors define an epigenetic strategy to therapeutically restore the ZBTB7A tumor suppressor function, demonstrating potent anti-leukemic activity in hematologic malignancies.

Firstly, to identify genes that suppress leukemia differentiation, the authors performed phenotypic CRISPR-Cas9 screens in two Cas9-expressing human AML cell lines. Differentiation was monitored via CD11b surface expression, detected using a fluorescently conjugated anti-CD11b antibody to identify AML blasts undergoing myeloid differentiation. Then, individual single guide RNA (sgRNA) abundance was analyzed by deep sequencing. In parallel, they conducted a positive-selection screen to identify genes whose loss promoted proliferation under differentiation-inducing conditions. Among the top-ranking genes identified was ZBTB7A, which encodes a POK domain that functions as a transcriptional repressor. To confirm their screen results, they individually transduced ZBTB7A sgRNAs (linked to a GFP reporter) into Cas9-expressing human AML cell lines and treated them with the differentiation-inducing agents. Results showed that ZBTB7A knockout (KO) cells exhibited reduced expression of the myeloid marker CD11b, and competition assays revealed an increased proportion of GFP+ cells in ZBTB7A KO versus control, indicating impaired differentiation. To examine Zbtb7a function in vivo, they used a Cre-inducible Cas9 mouse model and a Cre-expressing sgRNA vector encoding Zbtb7a sgRNA (which was transduced into hematopoietic stem and progenitor cells (HSPCs)) and transplanted into lethally irradiated host mice. Six weeks post-transplantation, they observed that Zbtb7a suppression led to a block in lineage-restricted progenitors and an expansion of HSC/MPP populations indicating that ZBTB7A loss suppresses AML differentiation. Conversely, cDNA of ZBTB7A (containing a GFP reporter) overexpression inhibited proliferation, promoted terminal differentiation and apoptosis, reduced colony formation, delayed leukemia progression, and prolonged survival of leukemia-bearing mice.

Mechanistically, chromatin immunoprecipitation sequencing (ChIP-seq) revealed that ZBTB7A binds to gene promoters and intronic regions. RNA sequencing (RNA-seq) analysis in Cas9-expressing AML cells showed transcriptional changes upon ZBTB7A loss. As expected for a transcriptional repressor, ZBTB7A KO led to widespread gene depression. Gene ontology (GO) analysis of the top upregulated pathways revealed key inflammatory responses and inflammatory cytokines and chemokines. Given their roles in shaping the inflammatory milieu of myeloid malignancies, they quantified TNF, IL-1β, MPO, and MMP-9 protein levels by enzyme-linked immunosorbent assay (ELISA) in conditioned media from ZBTB7A KO versus control AML cells. Consistent with transcriptomic findings, ZBTB7A KO significantly increased secretion of these inflammatory proteins. Conversely, ectopic overexpression of ZBTB7A led to global transcriptional repression. According with the previous data, the most significantly downregulated pathways upon ZBTB7A overexpression included inflammatory pathways. Notably, ZBTB7A loss resulted in increased expression of TNF and TNFAIP3, while its overexpression repressed these genes. Besides, ChIP-seq confirmed ZBTB7A binding at TNF and TNFAIP3 promoters.

To address the clinical relevance of their findings, they analyzed ZBTB7A expression using previously published single-cell RNA-seq (scRNA-seq) data of bone marrow specimens from diagnosed adult and pediatric AML patients ⁵, along with healthy donors. Hematopoietic stem cells (HSCs) from AML patients exhibited the lowest ZBTB7A expression, which progressively increased during myeloid maturation, linking expression to differentiation state. Analysis of an independent scRNA-seq cohort of patients ⁶ confirmed this pattern, showing the lowest ZBTB7A levels in HSC and early progenitor populations relative to mature myeloid cells. Given that high relapse rates and poor clinical outcomes in AML are largely driven by Leukemia Stem Cells (LSCs) ⁷, they examined the relationship between ZBTB7A expression and LSCs across three large AML cohorts ⁸⁹¹⁰. Interestingly, ZBTB7A ranked among the top 2% of genes with strongest negative correlation to LSC17, indicating reduced expression is linked to stem-like states. Finally, analysis of AML patients from a different cohort ¹¹ showed that lower ZBTB7A levels correlated with significantly worse overall survival. To explore the potential of reactivating ZBTB7A expression in AML, they developed a CRISPR-based screening approach to identify upstream regulators of TSGs. They designed fluorescent in situ hybridization (FISH) probes conjugated with a fluorophore (Alexa Fluor-647) to target the coding region of the ZBTB7A transcript. Next, they transduced AML Cas9 cells with a human genome-wide CRISPR/Cas9 knockout library ¹². Ten days post-transduction, AML cells were fixed and incubated with the ZBTB7A FISH probe and sorted based on ZBTB7A expression levels. Deep sequencing was then performed and upstream transcriptional and epigenetic regulators whose ablation increased ZBTB7A expression in AML cells were prioritized. Among the top candidates KDM4B was identified and, interestingly, ENCODE ChIP-seq analysis showed that KDM4A and KDM4B histone demethylases were bound to the ZBTB7A promoter.

To restore ZBTB7A activity they used a pharmacological approach through KDM4 inhibition using QC6352 inhibitor (QC6352 is a potent pan-KDM4 inhibitor which effectively targeted KDM4A-C) ¹³. Treatment of human AML cell lines with QC6352 led to an approximately 3-fold increase in ZBTB7A mRNA levels. Moreover, QC6352 also induced a dose-dependent increase in ZBTB7A protein levels. Then, they conducted a time-course experiment and treated AML line cells with QC6352 and found that induced myeloid differentiation followed by apoptosis. Moreover, KDM4 inhibition induced robust anti-leukemic activity and terminal differentiation across a broad range of AML cell lines and PDX models.

To assess the in vivo efficacy of KDM4 inhibition, they tested QC6352 in both AML cell line and PDX models. AML cells expressing firefly luciferase (an enzyme that emits photons after degradation of luciferine compound) were transplanted into immune-deficient recipient mice. Upon disease establishment, mice were treated with QC6352. Notably, bioluminescent (photon) quantification and imaging of leukemia-bearing animals treated with QC6352 showed marked delay in AML burden. Consistent with in vitro results, QC6352 induced myeloid differentiation and apoptosis in vivo at day 10, reduced AML burden, and significantly extended survival. They further extended these observations in two AML PDX models. Following engraftment in NSG-SGM3 mice, they were treated with QC6352 or vehicle. Remarkably, QC6352 treatment significantly reduced human AML cells in blood and bone marrow and decreased spleen infiltration and weight. They also observed that QC6352 treatment increased myeloid differentiation and apoptosis and significantly prolonged the survival of leukemia-bearing PDX mice compared to vehicle controls.

In summary, this study identifies the transcriptional repressor ZBTB7A as a TSG downregulated in AML, associated with poor survival outcomes. To uncover therapeutic strategies, they developed a methodology pinpointing the KDM4 family of histone demethylases as a vulnerability to restore ZBTB7A function. Thus, pharmacological inhibition of KDM4 upregulated ZBTB7A expression, promoted terminal differentiation of leukemic cells and prolonged mice survival. These findings reveal critical regulatory mechanisms of ZBTB7A and support epigenetic therapy as a promising strategy to reactivate its tumor suppressor function in hematologic cancers.

The post Epigenetic reactivation of a tumor suppressor program in AML appeared first on Mapping Ignorance.

A team of researchers explores now ¹ an unusual version of that phenomenon using two materials: a very thin layer of silicon carbide sitting on a semiconductor called indium arsenide. Silicon carbide has a characteristic vibration, called a phonon, in which neighbouring atoms move against each other. Indium arsenide, when it contains enough mobile electric charges, supports a surface electromagnetic wave called a surface plasmon polariton, which travels along its surface. The goal was to bring the energy of these two excitations into the same range so that they would interact strongly, and even ultrastrongly.

To do this, the researchers used a clever trick to tune the semiconductor on demand. A very short pulse of near-infrared light strikes the indium arsenide and temporarily knocks loose additional mobile electrons, briefly raising their concentration to roughly 10¹⁸ to 10¹⁹ per cubic centimetre. This shift changes the frequency of the surface plasmon polariton. By adjusting the delay between this pulse and a second, infrared probe pulse, the surface wave could be tuned until its frequency approached that of the silicon-carbide vibration.

The interaction itself was observed with a nanoscale infrared microscope. A metal-coated atomic-force-microscope tip, its very apex only tens of nanometres across, concentrates the infrared electric field into a spot far smaller than the wavelength of the light itself. This trick lets the instrument detect surface waves that are far more tightly confined than ordinary infrared measurements can reach, overcoming a long-standing barrier: free-space light and tightly confined surface waves normally cannot exchange energy efficiently, so ordinary spectroscopy cannot probe such confined waves directly. The scattered infrared light is analysed for both its strength and its phase, revealing how the excitations behave across both frequency and space.

Ultrastrong-coupling regime

The clearest sign of strong coupling is that a single resonance does not simply shift when two excitations meet; it splits into two new resonances, one above and one below the original frequency. Here, the measured splitting reached about 180 wavenumbers, more than 20 percent of the silicon-carbide vibration frequency, placing the system solidly in the ultrastrong-coupling regime. Calculations reproduced the observed splitting and confirmed that the new states are indeed a genuine mixture of the semiconductor surface wave and the silicon-carbide vibration.

The most surprising result appeared when the coupling was examined not just as a function of frequency but of the wave’s momentum, a quantity related to how tightly a wave is confined in space: the more confined the wave, the higher its momentum. In ordinary strong or ultrastrong coupling, two excitations mix substantially only over a narrow range near the point where their energies coincide. Here, however, the surface plasmon polariton’s frequency barely changes across a wide range of momenta once it becomes tightly confined against the surface, essentially flattening out. Because of this “flat” behaviour, the wave stays close enough in energy to the silicon-carbide vibration to remain strongly mixed across an unusually broad range of momenta, rather than only at a single, sharply defined point. The resulting hybrid states are described as having a flat band.

This spatial behaviour was confirmed directly by scanning the nanoscale probe across the edge of the silicon-carbide layer. Hybrid surface waves launched at the tip travel outward, reflect off the edge, and interfere with the incoming field, creating a pattern of intensity maxima and minima. Analysing this interference pattern reveals both the wavelength and the direction of travel of the hybrid waves. The resulting dispersion, that is, how the frequency of the wave changes with its momentum, closely matched theoretical predictions, providing independent confirmation of the unusually broad coupling.

There is a trade-off. A flat dispersion means these hybrid excitations travel with a very small group velocity, so they cannot efficiently carry energy over long distances. Even so, because many more momentum states can now participate in the coupling than in conventional systems, a much larger family of hybrid states becomes accessible.

The study also predicts, though does not yet demonstrate, that a similar flat-band effect could occur when molecular vibrations couple to low-loss surface phonon polaritons: electromagnetic surface waves that ride along vibrations in a polar crystal, related to silicon carbide’s own phonon but propagating along the surface rather than confined within a thin layer. Coupling light strongly to molecular vibrations, known as vibrational strong coupling, has attracted attention for its potential to influence chemical reactivity, although exactly how and why this happens is still being worked out. The significance of this work is therefore not a ready-made chemical technology, but a demonstrated way to generate and observe a much larger family of strongly hybridized light-matter states, extending the reach of strong coupling into a regime where the usual boundary between light and matter begins to blur.

Author: César Tomé López is a science writer and the editor of Mapping Ignorance

Disclaimer: Parts of this article may have been copied verbatim or almost verbatim from the referenced research paper/s.

The post Tuning a crystal until light and sound fuse appeared first on Mapping Ignorance.