The central question was simple: when molecules are compressed, which parts of them change the most? The answer turned out to depend on three main factors: how stiff a bond is, how much the molecular volume changes when atoms move, and whether the structure is a stable molecule or a transition state.

Microscopic effects of high pressure

To investigate this, the researchers used a computational method called the Extreme-Pressure Polarizable Continuum Model, or XP-PCM. Instead of treating pressure as a direct mechanical push on atoms, the model surrounds a molecule with a compressed environment that mimics matter under pressure. As that space shrinks, the molecule’s own electrons are repelled more and more strongly by those of the surrounding medium, faithfully reproducing the microscopic effects of high pressure.

The calculations showed that strong covalent bonds resist pressure remarkably well. In methane and benzene, ordinary carbon–hydrogen and carbon–carbon bonds shortened by less than 0.01 ångström between 0 and 5 GPa. An ångström is one ten-billionth of a metre, so these changes are extremely small. Strong bonds behave like stiff springs: pressure can compress them only slightly.

Weaker interactions responded much more dramatically. In diborane, which contains unusual electron-deficient bonds, the weaker bridging bonds compressed more than ordinary boron–hydrogen bonds. Even larger changes appeared in systems held together mainly by nonbonded interactions, where there is no strong chemical bond resisting compression. In molecules such as cyclophane, where two benzene rings sit close together without forming direct covalent bonds, the inter-ring distance shrank by about 0.05 to 0.07 ångström over the same pressure range. Weak interactions are therefore far more compressible than strong covalent bonds.

Metal-containing molecules displayed intermediate behaviour. In ferrocene, an iron atom sits between two cyclopentadienyl rings like a sandwich filling. Under pressure, the iron–carbon distances shortened noticeably more than ordinary covalent bonds but less than weak nonbonded contacts. The calculations also reproduced an experimentally observed pressure-induced structural change in ferrocene, showing that the computational approach captures real molecular behaviour accurately.

One of the most interesting aspects of the study involved transition states. Transition states are unstable arrangements of atoms located at the top of the energy barrier separating reactants and products. Because reactions pass through them only briefly, they are difficult to study experimentally. Pressure can change their geometries in surprising ways.

Why would compression stretch a bond?

In most stable molecules, pressure shortens bonds. Yet some transition states behaved in the opposite manner. In the transition state of the Diels–Alder reaction between butadiene and ethene, the forming carbon–carbon bonds actually became longer under pressure. At first this seems counterintuitive: why would compression stretch a bond?

The explanation comes from the shape of the reaction’s energy landscape. Pressure lowers the reaction barrier, shifting the transition state toward an earlier stage of the reaction. In that earlier structure, the reacting molecules are farther apart, so the forming bonds are longer. Rather than simply squeezing every bond equally, pressure changes where the transition state sits along the reaction pathway.

Other reactions showed different behaviour. In some electrocyclic reactions, pressure lengthened key transition-state distances, while in others it shortened them. The outcome depended on the precise atomic motions involved in the reaction. If the important motion mainly involved bond stretching, pressure tended to favour elongation in the transition state. If rotational motions dominated instead, ordinary bond compression could outweigh the stretching effect.

The study also examined hydrogen-transfer reactions and the Cope rearrangement, a classic organic reaction involving simultaneous bond breaking and bond formation. In these cases, pressure significantly reduced certain long, weak transition-state distances. For the Cope rearrangement, the compression was especially large — up to 0.2 ångström at 5 GPa — comparable to the structural change caused by adding certain chemical substituents to the molecule.

A key feature of the work was its mode-by-mode analysis. Molecules vibrate in characteristic patterns called normal modes. By examining how pressure affects each vibrational mode separately, the calculations identified exactly which molecular motions control structural changes under compression. In ferrocene, for instance, the mode that brings the two rings closer to the iron proved far more sensitive to pressure than the stiffer carbon–hydrogen stretching. Soft vibrations associated with weak interactions were most sensitive to pressure overall, while stiff stretching modes of strong covalent bonds changed little.

Overall, the study established a clear hierarchy. Strong covalent bonds are least affected by pressure, metal–ligand bonds are more compressible, and weak nonbonded interactions are the most sensitive of all. Transition states add another level of complexity, because pressure can sometimes push them toward structures with unexpectedly longer bonds.

These results offer more than a catalogue of compressed molecules. They provide a framework for predicting how chemical reactions and molecular structures behave in extreme environments, from deep inside planets to high-pressure industrial chemistry. Pressure is not merely a tool for squeezing matter closer together. It can reshape energy landscapes, redirect reaction pathways, and reveal unexpected features of chemical bonding itself.

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 Why would pressure stretch a chemical bond? appeared first on Mapping Ignorance.

 

Imagine trying to build a house without a blueprint, find a shortcut through an unfamiliar city without a map, or govern a large organisation with no leaders and no meetings.

It sounds impossible. Yet tiny-brained ants, working without leaders or blueprints, have been solving problems like these for millions of years – and no, the queen isn’t the boss telling them what to do.

By almost any measure, ants are a wildly successful group of animals – there’s an estimated 20 quadrillion of them on Earth and they thrive on every continent but Antarctica.

How have these minuscule animals managed to take over the world (and our kitchens)? The answer is teamwork.

Bustling colonies

Ants are social animals that live in colonies ranging from a few individuals to vast continent-spanning supercolonies containing billions of ants.

Bustling ant colonies display many of the features we associate with human societies, including:

• transportation networks
• collective care of the young
• food systems (including agriculture in some species)
• health care for injured nestmates.

In humans, this level of social complexity usually involves clear governance hierarchies, with leaders and middle managers directing our activities.

But ants don’t work that way. So who is in charge in an ant colony?

The answer is simple: no one.

The queen isn’t in charge

Ant colonies are a classic example of a self-organised system, where complex behaviour emerges from the combined actions of many ants. Each follow relatively simple rules while communicating and interacting with each other.

The human brain works in a similar way: individual neurons have simple behaviours and cannot think on their own, but together they give rise to the full range of human thought and behaviour.

The queen, whom many people assume is in charge, has little involvement in decision-making or leadership.

Instead, her role is to maintain the colony’s workforce by producing new ants.

In some ant species, workers will even kill their queens under particular conditions, such as declining productivity!

By working together, ant colonies are capable of complex behaviours and problem-solving skills far exceeding the abilities of an individual ant.

For example, some ant species run sophisticated transportation networks linking their colony to many food sources.

When a foraging worker finds a good source of food, such as some crumbs in your kitchen, she lays down drops of attractive chemicals called “pheromones” as she walks home.

Other ants in the colony are attracted to the trail, reinforcing it with more pheromones as they go. As a result, the colony can rapidly deploy large numbers of workers to quickly collect food.

While an individual ant is only aware of the foods she herself has visited, the trail network allows the colony as a whole to be “aware” of many foods.

Should a food source disappear or decline in quality, the colony can quickly refocus its efforts.

Ants can also optimise their trail networks by finding shortcuts.

Since pheromone trails evaporate over time, shorter paths that are traversed more quickly get reinforced more often. Longer paths, by contrast, receive less traffic and get reinforced less often, which in turn causes the pheromone trail to fade and become less attractive.

This simple feedback loop allows the colony to “discover” shorter routes that take less time to traverse while eliminating longer routes.

The resulting transportation network can be remarkably efficient.

Remarkable architects

Nest construction is another impressive example of the power of self-organisation.

Ant nests can be vast and intricately structured, with chambers for raising the young, food storage, and waste.

Yet no ant has a blueprint for the final nest design, nor is a boss ant in charge of directing construction activities.

Instead, ants use simple rules to create their remarkable nest architecture.

For example, in the black garden ant Lasius niger, nest building ants excavate soil and form it into small pellets.

These pellets carry chemical cues making other ants more likely to deposit their own pellets nearby.

Over time, this leads to the formation of structures such as pillars, walls, and eventually roofs, without any ant understanding the overall design.

This process, where individuals respond to cues left behind by other individuals, is called “stigmergy” and it underpins the construction of other insect-built structures such as termite mounds and honeycomb.

More humans, more problems – but not so for ants

The use of simple behavioural rules enables ants to coordinate remarkably effectively as a group.

In a study where groups were tasked with moving a T-shaped object through a tight space, human performance did not improve with group size.

When participants were instructed not to speak, performance actually declined as groups got bigger.

Similarly, it has long been known that as human group size increases, the performance of individual team members tends to decrease, a phenomenon known as the Ringelmann effect.

Ants, by contrast, showed the opposite pattern: as group size increased, their performance actually improved.

So next time you see a line of ants marching around your house, resist the urge to spray or whack them away.

Instead, take a moment to appreciate these tiny masters of teamwork.

 

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

The post In an ant colony, the queen isn’t in charge. So who is? appeared first on Mapping Ignorance.

“The hot springs along the Kafue rift of Zambia have helium isotope signatures which indicate that the springs have a direct connection with Earth’s mantle, which lies between 40 and 160km below Earth’s surface,” said Prof Mike Daly of the University of Oxford, an author of the article. “This fluid connection is evidence that the fault boundary of the Kafue Rift is active and therefore the Southwest African Rift Zone is too—and may be an early indication of the break-up of sub-Saharan Africa.”

The bubbling gun

The Kafue Rift is part of a 2,500 km long zone of rifts that runs from Tanzania to Namibia and may reach the mid-Atlantic ridge. The scientists’ attention was drawn to it by topography that suggested a possible new rift, as well as high levels of geothermal anomalies and hot springs.

But to confirm a new rift, scientists would need to show that it had broken through Earth’s crust: evidence that fluids had escaped from the liquid mantle to the surface.

“A rift is a large break in Earth’s crust that creates subsidence and associated elastic uplift,” said Daly. “A rift may become a plate boundary, but commonly a rift’s activity ceases before the point of lithospheric break-up and plate boundary formation.”

The scientists visited eight geothermal wells and springs across Zambia: six in the suspected rift zone, and two outside it. They took samples of gas from freely bubbling water, and analyzed these in the laboratory to identify the isotopes of each element present.

Isotopes are different forms of an element, which are present in different proportions in the crust and in the mantle. So, by testing the isotopes present in the gas, the scientists could detect the presence of gas derived from mantle fluids at the surface. They compared these to readings taken from the East African Rift System, an ancient, well-established rift.

Earth on the move

The scientists found that the gas from the Kafue Rift, but not the gas from the springs outside the rift, contained a ratio of helium isotopes comparable to samples taken from the East African Rift System.

The helium couldn’t have come from the atmosphere, because the ratios of helium isotopes weren’t consistent with those found in the air, or just from the crust, because there was too much of the mantle-sourced helium isotope present for that.

The Kafue Rift samples also contained a proportion of carbon dioxide consistent with carbon dioxide found in mantle fluids. Helium isotopes provide a signal of early-stage rifting: Using the East African Rift System as a model, scientists predict that with time, carbon dioxide will become more prominent as volcanic centers develop.

The discovery that the Kafue Rift is active could have important economic implications. Early-stage rifts can provide geothermal energy and access to helium and hydrogen where they are not diluted by the volcanic gases. However, it could have even more significant implications for the future shape of Africa.

“Many of the features of the Great Rift Valley of Kenya offer compelling reasons why East Africa should ultimately become a line of major continental break-up,” said Daly.

“But the rate of rifting of the East African Rift System is slow. On almost all sides of Africa there are mid-ocean ridges tending to inhibit east-west or north-south extension, so break-up and spreading does seem to struggle to establish itself.

“The Southwestern African Rift System could be an alternative. It has the required rift-related features, and regional basement fabrics—inherent weaknesses in the crust—favorably aligned to the surrounding mid-ocean ridges and continental geomorphology. This relationship may offer a much lower strength threshold for continental break-up.”

“However, this study is based on helium analyses from one general area in the Southwest African Rift System, which is thousands of kilometers long,” cautioned Daly. “This early study is being followed by more extensive studies, the next step of which will be completed this year.”

The post New tectonic plate boundary could be forming in Zambia appeared first on Mapping Ignorance.

Two years ago, in March 2024, I wrote in this blog about a decision that surprised many: the rejection of the Anthropocene as a formal geological epoch. After fifteen years of work by an international team of experts, the proposal to declare that we are now living in a new chapter of Earth’s history—one defined by human impact—had been voted down by the Subcommission on Quaternary Stratigraphy.

At the time the debate was intense. Accusations of procedural irregularities circulated. The historian Naomi Oreskes described the vote as “a sham”. The International Union of Geological Sciences had to intervene to confirm that the process had followed its rules. Amid the controversy, one question stood out: should the Anthropocene be treated as a formal “epoch” with a single, precise start date, or as an ongoing “event” that unfolded at different moments across the planet?

Two years later, in 2026, the dust has largely settled. The official International Chronostratigraphic Chart has not changed: we remain in the Holocene. Yet new articles, conference sessions and books about the episode continue to appear. With some distance, we can now ask a different set of questions: what does this episode reveal about how science actually works? And why should anyone who is not a geologist care?

The clash that wouldn’t go away

The popular story of the Anthropocene is comforting: a brilliant scientist has a sudden insight, coins a powerful term, evidence accumulates, consensus emerges, and a new epoch is added to the geological timescale.

Reality proved more complicated. What unfolded was a deep clash between two scientific cultures, each shaped by its own history, values and standards of evidence.

One side was Earth System Science, a field that took shape in the 1980s around the idea that the planet operates as a single, interconnected system—oceans, atmosphere, biosphere and human activity all influencing each other. Its roots reach back to Cold War-era projects like the International Geophysical Year (1957–58), were strengthened by NASA’s search for a post-Apollo mission, and were institutionalised through large international programmes. Earth system scientists think in terms of global datasets, computer models and planetary boundaries. For them, the Anthropocene was evident: the hockey-stick curve of CO₂, plastic everywhere in the oceans, accelerating biodiversity loss. The Earth system had clearly shifted.

On the other side stood traditional stratigraphy, a discipline with nineteenth-century origins. Stratigraphers act as the custodians of deep time, maintaining a single, consistent global timeline so geologists anywhere can refer to the same moment in Earth history. To recognise a new epoch they demand a “golden spike” (a Global Boundary Stratotype Section and Point, or GSSP): a specific, accessible rock section where a change is recorded that is both global and synchronous (happening at the same time everywhere). These rules, refined over decades, prioritise stability, precision and international agreement.

The Anthropocene Working Group tried to bridge both worlds. It was unusually interdisciplinary—stratigraphers, Earth system scientists, historians and even a legal expert. They proposed a mid-twentieth-century start date marked by radionuclides from nuclear weapons tests and selected Crawford Lake in Canada as the reference site. For Earth system scientists this was logical: the planetary transformation was real, and the radioactive layer was simply one readable signal among many.

The stratigraphers, however, judged the proposal against narrower criteria: did it satisfy the established rules for a GSSP? Their answer was no.

Whose objectivity?

What struck me then—and still does—is how each side accused the other of straying from science.

Naomi Oreskes argued that by refusing to recognise what “we all can now see”, the stratigraphers risked damaging the credibility of science itself.

From the stratigraphers’ viewpoint the situation looked different. Years earlier, two of them had warned that the drive to formalise the Anthropocene was driven more by political and public resonance than by stratigraphic evidence. Visibility in policy circles, emotional weight, media attention—these factors, they felt, were influencing the process beyond the technical standards that had governed their field for generations.

Both groups claimed to defend objective science. Both believed the other was allowing external values to intrude. And both, I believe, were partly right—and partly missing the deeper point.

The philosopher Helen Longino offers a useful lens here. Objectivity, she argues, is not the absence of values (an impossible goal); it is the presence of procedures that expose ideas to criticism from diverse perspectives. The stratigraphers had their procedures: rigorous GSSP criteria, voting protocols, long-established consensus mechanisms. The Earth system scientists had theirs: interdisciplinary teamwork, attention to planetary-scale dynamics, concern for societal relevance.

The conflict was not science versus politics, nor pure objectivity versus bias. It was a clash between two legitimate but differently organised ways of producing trustworthy knowledge. The question “Is the Anthropocene real?” only makes sense once we add: “Real according to whose standards and practices?”

A forgotten alternative

In my earlier post I noted the idea—defended by some scholars—that the Anthropocene might be better understood as an ongoing geological event (like the spread of life onto land) rather than a sharply bounded epoch.

There is another angle worth recovering. In 1873 the Italian geologist Antonio Stoppani proposed ¹ an “Anthropozoic era”. He is occasionally cited as an early precursor to the modern concept, slotted into a linear story that leads to Paul Crutzen. But when we read Stoppani on his own terms, he offers something more valuable: a different way of imagining human impact.

Stoppani’s markers were not global geochemical spikes or planetary system shifts. They were material and archaeological: human bones, tools, buildings, settlements preserved in terrestrial deposits—caves, lake sediments, river deltas. He drew evidence mostly from Europe and openly acknowledged that extending the idea worldwide would require far more data. His “era” encompassed the entire span of human presence, not merely the decades since 1950.

Stoppani did not have our satellites, ice cores or global monitoring networks. Yet precisely because his imagination operated with different priorities and evidence, he serves as what the philosopher Hasok Chang calls a “diagnostic contrast”. Comparing his approach with ours helps us see the specific choices embedded in the contemporary Anthropocene concept.

Why do we privilege the global scale above all others? Why do we focus so heavily on predicting future trajectories rather than documenting the material traces of the past? Why do we anchor the concept in a thin layer of radioactive particles rather than in the dense accumulation of cities, mines, fields and landfills? These are not merely historical curiosities. They influence how we frame today’s environmental crisis and which responses we consider plausible.

What the debate tells us, two years on

In 2026 the stratigraphic rejection stands. No proposal to revisit the Anthropocene epoch has gained traction within the International Commission on Stratigraphy. Yet the concept itself has not faded—quite the opposite. In the social sciences, humanities and policy discussions it continues to thrive. Terms like Capitalocene (centring economic systems) and Plantationocene (highlighting colonial agriculture and slavery) have gained ground, each drawing attention to different drivers of planetary change.

The formal rejection may have done the concept a favour by releasing it from the strict requirements of one discipline. Freed from the need for a single golden spike, the Anthropocene has become a more flexible frame for thinking about humanity’s role on Earth.

The larger lesson, I think, concerns scientific authority itself. Authority is not a monolithic property that science either possesses or lacks. It is negotiated, distributed and sometimes contested across communities. The stratigraphers preserved control over the official geological timescale, but at the price of seeming disconnected from a powerful idea that had already entered public culture. The Earth system scientists saw their framing continue to shape research and debate even without formal recognition.

For everyone else—researchers in other fields, students, policymakers, concerned citizens—the episode reminds us that scientific categories are never simply “discovered” in nature. They are constructed through evidence, yes, but also through values, disciplinary traditions, institutional histories and contingent choices.

Recognising this does not weaken science. It makes the enterprise more transparent, more interesting—and more human.

The Anthropocene may never appear on the official chart. But it will keep shaping how we understand our place on the planet. And understanding why that is so tells us as much about science as it does about the Earth.

Acknowledgements: José Luis Granados Mateo’s research is supported by funding from the Basque Government’s Postdoctoral Research Program.

Further reading:

• Chang, H. (2012). Is Water H₂O? Evidence, Realism and Pluralism. Springer.
• Longino, H. (1990). Science as Social Knowledge. Princeton University Press.

The post What the Anthropocene rejection reveals about scientific cultures appeared first on Mapping Ignorance.

A recent study ¹ examined this phenomenon in a family of layered crystals called NbOX₂, where X can be chlorine, bromine, or iodine. These materials already attracted attention because they combine several unusual properties: they are ferroelectric, meaning they possess an electrically switchable internal polarization, and they interact strongly with light in ways useful for photonic devices. The new work shows that they may also provide a controllable platform for chiral quantum matter.

The central question is how a crystal that initially has no handedness can spontaneously become left-handed or right-handed. The answer lies in the way atoms arrange themselves inside the material.

An achiral intermediate structure

At high symmetry, the crystal structure is achiral. In this state, the arrangement of atoms looks identical to its mirror image. However, calculations show that this structure is unstable. Tiny atomic vibrations naturally push the system toward a lower-energy arrangement. Instead of remaining perfectly symmetric, the atoms shift slightly and reorganize into a chiral structure with a definite handedness.

This process resembles balancing a pencil vertically on its tip. The upright position is symmetric but unstable. A tiny disturbance causes the pencil to fall in one direction or the other. In the crystal, the “falling direction” determines whether the material becomes left-handed or right-handed.

The study identifies an important intermediate stage between the symmetric and chiral phases. This intermediate structure is still achiral, but it has a very shallow energy well, making it susceptible to stabilization by external perturbations such as pressure (predicted around 81 kbar for NbOCl₂), thermal or quantum anharmonic fluctuations, or optical excitation. That makes the material unusually tunable.

The researchers mapped the crystal’s energy landscape using first-principles quantum calculations. The resulting landscape resembles a “Mexican hat” shape often encountered in symmetry-breaking physics. At the center sits the unstable symmetric state. Around the rim lie several low-energy states corresponding to different distortions of the crystal. Some distortions remain achiral, while others produce chirality.

Switchable chirality

An especially important result is that the energy barrier separating left-handed and right-handed structures is extremely small. This means the crystal could potentially switch handedness relatively easily. The study proposes a mechanism for controlling this process with an external electric field.

The electric field slightly favors one chiral arrangement over the other by breaking the mirror symmetry between them. Temperature or pressure, through anharmonic atomic fluctuations, effectively lower the energy barrier between the two enantiomers, facilitating the transition toward the preferred one. Once the field is removed, the selected chiral state can remain stable. In effect, the crystal behaves somewhat like a compass needle aligning with a magnetic field, except the field controls chirality instead of direction.

An obstructed atomic limit

Beyond chirality itself, the work also connects these materials to topological physics. In condensed matter physics, topology describes properties that remain robust even when a material is deformed or disturbed. Topological materials often host special electronic states on their surfaces or edges that are protected against defects.

The calculations reveal that the chiral phase of NbOX₂ contains an unusual electronic structure called an obstructed atomic limit. The electrons behave as though their charge is concentrated in regions that do not coincide directly with the atomic positions. This hidden organization can generate topological surface states under certain crystal cleavages; though notably not along the surface most easily exposed in the laboratory, which does not intersect the relevant charge centers.

Flat bands

The material also develops very flat electronic bands just below the Fermi level, the threshold energy separating occupied from unoccupied electron states. Flat bands are important because electrons in them have nearly zero kinetic energy, allowing their mutual repulsion to dominate and drive collective quantum behavior. Strong interactions can produce exotic quantum phenomena such as correlated insulating states or unconventional superconductivity. Similar ideas have become central in research on twisted graphene and other engineered quantum materials.

The study therefore connects several modern themes in condensed matter physics at once: spontaneous symmetry breaking, ferroelectricity, chirality, topology, and strongly interacting electrons. What makes these crystals particularly interesting is that all these effects appear in a single material family that may be experimentally controllable through electric fields, temperature, and pressure.

More broadly, the work demonstrates how chirality can emerge not merely as a static geometric feature, but as a dynamic property that can potentially be switched and manipulated. If such control becomes practical, chiral layered materials could eventually contribute to optical technologies, spin-based electronics, and future quantum devices where topology and chirality work together.

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 Switchable chirality in layered materials appeared first on Mapping Ignorance.

 

As Homer tells us, Odysseus made an epic journey, against the odds, from Troy to his home in Ithaca. He visited many lands, but mostly dwelt with the nymph Calypso on her island.

We can imagine that his wife, Penelope, would have asked him about that particular time. Odysseus might have replied, “It was nothing. In fact, it was less than nothing. Negative five years I dwelt with Calypso. How else could I have arrived home after only ten years? If you don’t believe me, ask her.”

Quantum particles, it turns out, are just as wily as Odysseus, as we have shown in an experiment published in Physical Review Letters. Not only can their arrival time suggest that they dwelt with other particles for a negative amount of time, but if one asks those other particles, they will corroborate the story.

Photons dwelling with atoms

Our experiment used photons – quantum particles of light – and the against-the-odds journey they must undertake to pass straight through a cloud of rubidium atoms.

These atoms have a “resonance” with the photons, meaning the energy of the photon can be transferred temporarily to the atoms as an atomic excitation. This allows the photon to “dwell” in the atomic cloud for a time before being released.

For this resonance to be effective, the photon must have a well-defined energy, matching the amount of energy required to put a rubidium atom into an excited state.

But, by a form of Heisenberg’s famous uncertainty principle, if the energy of the photon is well defined then its timing must be uncertain: the pulse of light the photon occupies must have a long duration. This means we can’t know exactly when the photon enters the cloud, but we can know on average when it enters.

If a photon like this is fired into the cloud, the most likely outcome is that its energy will be transferred to the atoms, and then re-emitted as a photon travelling in a random direction. In such cases, the photon is scattered, and fails to arrive at its Ithaca.

Photon arrival times

But if the photon does make it straight through, a strange thing happens. Based on the average time when the photon enters the cloud, one can calculate the expected average time it would arrive at the far side of the cloud, assuming it travels at the speed of light (as photons usually do).

What one finds is that the photon actually arrives far earlier than that. In fact, it arrives so early it appears to have spent a negative amount of time inside the cloud – to exit, on average, before it enters.

This effect has been known for decades and was observed in a 1993 experiment. But physicists had mostly decided not to take this negative time seriously.

That’s because it can be explained by saying that only the very front of the long-duration pulse makes it straight through the atomic cloud, while the rest is scattered. This leads to a successful (non-scattered) photon arriving earlier than would be naively expected.

Asking the atoms

However, Aephraim Steinberg, one of the authors of that 1993 paper, was not so quick to accept this dismissal of the negative time as an artefact. In his laboratory at the University of Toronto, he wanted to find out what happened if one queried the rubidium atoms in the cloud to find out how long the photon had spent dwelling among them as an excitation. After an initial experiment with inconclusive results, he asked me, as a quantum theorist, for help in working out what to expect.

When we talk of querying the atoms, what this means in practice is continuously making a measurement on the atoms while the photon is passing through the cloud, to probe whether the photon’s energy is currently dwelling there. But there is a subtlety here: measurements in quantum physics inevitably disturb the system being measured.

If we were to make a precise measurement of whether the photon is dwelling in the atoms, at each instant of time, we would prevent the atoms from interacting with the photon. It is as if, merely by watching Calypso closely, we would stop her getting her hands on Odysseus (or vice versa). This is the well known quantum Zeno effect, which would destroy the very phenomenon we want to study.

Negative time experiment

The solution is to make, instead, a very imprecise (but still very accurately calibrated) measurement. That is the price paid to keep the disturbance negligible. Specifically, we fired a weak laser beam – unrelated to the single photon pulse – through the cloud of atoms, and measured small changes in the phase of the beam’s light to probe whether the atoms were excited.

Any single run of the experiment gives only a very rough indication of whether the photon dwelt in the atoms, but averaging millions of runs yields an accurate dwell time.

Amazingly, the result of this weak measurement of dwell time, when the photon goes straight through the cloud, exactly equals the negative time suggested by the photons’ average arrival time. Prior to our work, no-one suspected that these two times, measured in entirely different ways, would be equal.

Crucially, the negative value of the weakly measured dwell time cannot be explained by imagining that only the front of the photon’s pulse gets through, unlike the time inferred from the arrival time.

So what does this all mean? Is a time machine just around the corner?

Sadly, no. Our experiment is fully explained by standard physics.

But it does show that negative dwell time is not an artefact. However paradoxical it may seem, it has a directly measurable effect on the atomic cloud that the photon traverses. And it reminds us that there are still lands to discover on the odyssey that is quantum research.

 

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

The post ‘Negative time’ measured in the lab appeared first on Mapping Ignorance.

It’s the first language model designed for population genetics ¹,, said Andrew Kern, a computational biologist in the UO College of Arts and Sciences. The AI tool offers scientists a fast and flexible alternative to classical methods for reconstructing evolutionary history.

In practice, it can help researchers like Kern understand when disease-resistance genes emerged in a population, for example, or when species evolved key traits.

“Advances in generative AI and the architectures behind them are potentially useful to a number of fields outside a chatbot,” said Kern, an Evergreen professor of biology. “We’re borrowing strengths from the world of AI and applying them in this different context that’s largely been untapped.”

Training AI on the language of DNA

Genomes are often compared to a written language, with combinations of DNA’s four-letter alphabet—A, T, C and G—forming the basis for genes and chromosomes. Kern and his lab are most interested in what’s misspelled, which scientists call mutations: changes in DNA sequences, like swapped or missing letters, that accumulate over time as part of evolution.

Often harmless, mutations can be passed down from generation to generation, leaving a trail of breadcrumbs for tracing ancestral relationships.

Traditional methods based on math and statistics are the gold standard for translating mutations into ancestry. They’re difficult to beat in most cases, said Kevin Korfmann, lead author of the study and former postdoctoral researcher at the UO. But those classical probabilistic approaches can be slow and struggle with large or incomplete genomic datasets, he added.

DNA-reading AI

So, the researchers looked to AI to efficiently interpret the language of life by modifying a GPT-2 model, the older machine learning architecture behind ChatGPT. But instead of being trained on large volumes of English text, the language model was trained on simulations of genetic evolution across different species—including bacteria, rodents, mosquitoes and primates—to learn and recognize mutation patterns.

“We can’t repeat evolution, so one of the key workflows we have is developing simulations,” Korfmann said. “The simulations mimic evolutionary processes, and then we use the outcomes as training data for our deep learning models.”

In general, stretches of DNA with many mutations likely trace back to a distant common ancestor, whereas those with few mutations are likely to share a more recent ancestor. This helps explain why chimpanzees are considered humans’ closest living relatives, with similar DNA, while sea sponges are the most distant, having diverged genetically more than 700 million years ago.

Based on those mutation patterns and other biological principles, the AI model can predict when gene pairs last shared a common ancestor, known as the “coalescence time.”

Sidestepping data bottlenecks

In tests, the tool performed as well as state-of-the-art statistical methods, which was surprising to the research team.

“You never really know what’s going to work when you’re essentially borrowing techniques from a totally different world and applying them to a new problem,” Kern said. “But this was a case where things worked really well.”

The computer model was also dramatically faster. While traditional methods can take hours or even days to decode a single mosquito chromosome, the new approach can do it in minutes. That efficiency is especially beneficial for scientists handling large amounts of genetic sequence data.

“Compared to classical inferential approaches, the AI tool doesn’t have to reason about every mutation individually,” Korfmann said. “It just reads the patterns because all of the expensive statistical work was done up front, during training, which sidesteps the bottleneck.”

The model’s simulation-based training also enables scientists to use DNA datasets that are incomplete or missing genetic code—an issue Kern frequently faces when working with mosquito genetic databases for his research on malaria transmission.

That versatility comes at a crucial moment for malaria control, Kern said. For decades, insecticides have been a cornerstone for the prevention of malaria-spreading mosquitoes. But evolution, as Kern puts it, “did its thing.”

“Insecticide resistance is being observed in all of these mosquito populations today,” he said. “A major challenge in preventing the spread of malaria has been understanding the evolution of insecticide resistance. Now, we can go in with our AI model, ask how long ago these resistance genes arose in the population, and learn about the evolutionary history of this critical carrier of malaria.”

Looking ahead, Kern and Korfmann aim to advance the biological model beyond tracing shared ancestry between two lineages towards reconstructing full genealogical trees across multiple lineages. Some traditional methods can already do this, but Kern said they’d like to chase that goal from a machine-learning angle.

“There’s so much going on in the machine learning field that we haven’t applied yet in our field,” Korfmann said. “There’s tons of translational work to do to get these novel algorithms working in biology.”

The post DNA-reading AI reconstructs ancestry in minutes appeared first on Mapping Ignorance.

For decades, biology textbooks have repeated the same idea with almost ritual certainty: neural stem cells, the cells that fuel the birth of neurons and glia during development, live exclusively inside the brain and spinal cord. They are the architects of the central nervous system, and nowhere else. It has always seemed like an elegant rule of nature—clean, simple, and unbreakable. But science rarely stays still, and sometimes a single unexpected observation is enough to make a familiar landscape suddenly look foreign. A study ¹ done by Hang and collaborators published in Nature Cell Biology has done exactly that. A large international team has uncovered a population of neural stem cells located far beyond the borders of the central nervous system. They are not neural crest progenitors, nor reprogrammed cells, nor artefacts of culture. They are bona fide neuroepithelial‑derived stem cells, essentially the same type that builds the embryonic brain, quietly residing in tissues such as the lung, tail, dorsal root ganglia, and embryonic limb of the mouse. And that discovery forces us to rethink some of the most basic assumptions about how and where the nervous system forms.

The story begins with a serendipitous twist. The researchers were attempting to replicate a set of controversial (and later retracted!!) experiments claiming that exposing cells to a low‑pH shock could convert them into pluripotent stem cells. Their goal was simply to verify those claims. Instead, what they found were small clusters of cells that looked remarkably like neural stem cells, even though the tissue of origin was not the brain. The result did not match the original hypothesis, but it pointed to something far more intriguing. Perhaps these cells had not been created by the treatment; perhaps they had been there all along. That possibility changed the direction of the project. The team began systematically searching for these mysterious cells using transgenic mice in which neural stem cells glow under the microscope. Immediately, unexpected patterns began to appear. In the adult lung, for instance, a rare population of cells (peripheral neural stem cells abbreviated as “pNSCs”) lit up with the same fluorescent (SOX1-GFP⁺) markers that define neural stem cells in the hippocampus. When the researchers isolated them and grew them in culture, they behaved just like classical neural stem cells: they self‑renewed, formed stable colonies, and differentiated into neurons, astrocytes, and oligodendrocytes. Even more striking, when transplanted into the developing mouse brain, they integrated into the tissue, survived for weeks, and matured into functional neural cell types without forming tumors.

For decades, scientists have known that most neural‑like cells found outside the central nervous system come from the neural crest, a migratory population that gives rise to the peripheral nervous system. Yet these newly discovered cells didn’t fit that category. They lacked the molecular signatures of neural crest derivatives and, when the researchers used genetic tools to permanently mark all descendants of the neural crest, the new cells remained unmarked. That meant they had to come from a different origin. The breakthrough came with another lineage‑tracing experiment, this time using a mouse line that permanently labels neuroepithelial cells (these cells are the earliest progenitors of the central nervous system). The result was unambiguous: the peripheral stem cells were marked. They were descendants of neuroepithelial cells, just like the neural stem cells in the brain. The only difference was “geographical”. Somewhere early in development, a subset of these neuroepithelial progenitors seems to leave the neural tube and migrate into other tissues, where they settle, lie low, and remain as a hidden reservoir of neural stem cells.

To understand exactly who these cells are (and who they are not!) the researchers turned to single‑cell RNA sequencing, which allows thousands of individual cells to be profiled one by one. The comparison between brain‑derived neural stem cells and these peripheral stem cells was surprisingly tight. Their gene expression programs, transcriptional signatures, and developmental potentials overlapped almost perfectly. There were small differences, subtle enough to reflect the influence of their local environments rather than a distinct lineage identity, but the overall conclusion was crystal clear: these peripheral stem cells are genuine neural stem cells, not impostors and not a hybrid class. Moreover, the study found them in tissues like lung and tail, but not in others, such as the heart, where only fully differentiated neurons were detected. This selective presence suggests that their migration during development follows precise anatomical routes and that their persistence into postnatal life is controlled by finely tuned local niches. Many of these cells exist in a quiescent state (dormant, where cells only divide rarely) while others show signs of activation, hinting at an internal balance similar to what is seen in the neural stem cell compartments of the brain.

From the functional point of view these peripheral stem cells are far from decorative. The researchers showed that during embryonic and even postnatal development, these cells contribute to forming mature neurons in multiple tissues, including dorsal root ganglia and intestine. In some organs, their contribution is confined to early development, while in others it continues after birth. This suggests that the body maintains a much more distributed neurogenic potential than previously appreciated, opening the door to new concepts about how peripheral organs might regulate their own innervation. Thus, the implications of this discovery are broader and profound: If neural stem cells exist outside the central nervous system, why are they there? What purpose do they serve? Could they play a role in tissue repair after injury, or in adapting organ function to different physiological needs? Are they remnants of an ancient developmental program, or part of an underappreciated regulatory system? These questions remain open, but the answers could reshape our understanding of neural development and regeneration.

From a translational perspective, the idea is tantalizing, because access to neural stem cells within the brain is difficult and risky, however, peripheral tissues are far more accessible. If these newly identified stem cells can be harnessed, they could someday provide a source of neurons for repair therapies without the need for invasive brain surgery. This is, of course, speculative, and the present study does not claim any therapeutic application. But it establishes a conceptual foundation that future research will almost certainly build upon. What stands out most in the work of Han and collaborators is the elegance of its challenge to a long‑held belief. Scientific dogmas often persist not because they are unquestioned, but because no evidence has emerged to contradict them. Here, through careful experimentation and a willingness to follow unexpected results, the authors reveal that the nervous system’s developmental boundaries are more porous and more flexible than we thought. They uncover a kind of biological quiet zone: a dispersed, semi‑hidden network of neuroepithelial‑derived stem cells that sit outside the brain yet retain the full potential to act like neural stem cells.

This discovery does more than update the detail of developmental biology. It rewrites part of the map. It reminds us that the body is full of overlooked corners, of wanderers and settlers that defy the categories we impose on them. And it highlights a truth that recurs again and again in science: nature does not draw lines where we expect them. Sometimes, it leaves doors open, quietly, subtly, waiting for us to notice…

The post Neural stem cells discovered outside the mouse brain appeared first on Mapping Ignorance.

A thin layer of xenon on the surface of a topological insulator can do something surprising: it can create a moiré pattern that reshapes how electrons move. In a new study ¹, that idea was tested on two well-known topological insulators, Bi₂Se₃ and Bi₂Te₃, using a technique called angle-resolved photoemission spectroscopy, which maps the energy and momentum of electrons.

The key feature of a topological insulator is that its interior behaves like an insulator, while its surface supports special conducting states. These surface states form a Dirac cone, meaning their energy changes almost linearly with momentum. In a topological material, these states are protected by symmetry, so they cannot be altered in just any way. That is why the effect of a moiré pattern on them is especially interesting.

The xenon layer does not match the crystal underneath perfectly. That mismatch creates a long-wavelength pattern on the surface, like a repeating ripple. When the electronic structure is measured, the original topological surface band is copied into additional replicas shifted in momentum space. This is clear evidence that the moiré potential is affecting the surface electrons directly.

What makes this system unusual is that the copied bands do not behave the same way everywhere. At some crossings, the bands avoid each other and open small gaps. But at special symmetry points, the crossings remain gapless. That difference is exactly what topological protection predicts. In the Xe/ Bi₂Se₃ system, the largest observed gap is about 27 meV, and it appears near the K points of the moiré Brillouin zone.

These gaps matter because they create van Hove singularities, points where many electronic states crowd into a narrow energy range. When that happens, the density of states rises, and electronic interactions become stronger. One important interaction is electron-phonon coupling, the interaction between electrons and vibrations of the crystal lattice. Stronger coupling can help electrons form pairs, which is the basic ingredient of superconductivity.

The study finds that the electron-phonon coupling on the xenon-covered Bi2Se3 surface is stronger than on clean Bi₂Se₃. That does not mean superconductivity has been observed here. It means the surface has moved closer to conditions where superconductivity could become possible if the system is tuned further, for example by changing the carrier density or the moiré period.

This result shows that a simple, chemically inert layer of xenon can be used to engineer the electronic structure of a topological insulator surface in a controlled way. Instead of twisting two fragile crystals, the moiré pattern forms naturally through adsorption. That makes the approach cleaner and potentially easier to scale.

The broader significance is that moiré engineering is no longer limited to graphene-like systems. It can also be applied to topological insulators, where symmetry and topology add new rules to the physics. In this case, those rules produce protected crossings, gapped intersections, enhanced electronic interactions, and a possible route toward topological superconductivity in the future.

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 Moiré patterns on topological insulators, a new route toward topological superconductivity appeared first on Mapping Ignorance.

Recently, a cheerful 100-year-old message in a bottle was found on the south-west coast of Australia. In it, a world war one soldier proclaimed to be “as happy as Larry”.

If you’re a betting person, you probably wouldn’t expect great odds of this happening. A bottle cast into the ocean could end up absolutely anywhere.

If it floats to a remote location, there is little chance of somebody stumbling upon it. And if it lands somewhere more favourable where people could potentially find it, there are other issues. The message itself will deteriorate over time as light degrades it. If the bottle fills with water, it will sink and almost certainly never be found.

So, what are the chances of a message in a bottle being found and it being over 100? And what are your chances of finding this bottle?

Despite these many possibilities during a bottle’s lifetime, the probability we are after is a straightforward calculation. Just count up the number of bottles with messages that have been found and are over 100 years old, and divide by the number of messages that have been sent this way (assuming we know how many are sent):

Our diagram below shows a hypothetical situation where 20 bottles are sent in total, of which six are found (indicated in gold) and one of these is over 100 years old (indicated by the “100” stamp). So, one in 20 bottles are found and over 100 years old. (Note: This is only a hypothetical calculation, not the real data.)

Instead of calculating the probability directly, another way to do it is by breaking the problem into two parts: (A) a bottle with a message is found, and (B) the found bottle is over 100. These two probabilities can be calculated separately and multiplied together to get what we want:

This is known as the “multiplication rule” of probability, and we confirm from our hypothetical numbers that (6/20)×(1/6) = 1/20, as before.

Both approaches to calculating this probability are simple. However, the direct calculation requires knowing the total number of bottles sent out, which is very difficult to know in the real world.

The multiplication rule has the advantage that it breaks the calculation into two parts. We can tackle each separately, then bring the two results together to get the probability we want. This is useful in the real-world situation where we can draw information from different sources.

First, we’ll deal with the probability that a bottle with a message is found, irrespective of its age.

Experts from the Federal Maritime and Hydrographic Agency of Germany suggest a one in ten chance that a message in a bottle will be found. This aligns broadly with various historical “drift bottle” experiments, where oceanographers released large numbers of bottles to understand ocean currents.

For example, studies from the 1960s and ’70s in the North Atlantic Ocean led to recovery rates of 14% from the Gulf of Mexico, 8% from the Caribbean Sea and 7% from the northern Brazilian coast. A more recent and more northerly study (between Canada and Greenland) from the 2000s led to a 5% recovery rate.

We would expect the results to vary naturally from different experiments in different parts of the world. But to keep things simple, we will stick with 1/10 as the probability that a bottle with a message is found.

Now for the second piece of the calculation: of the bottles that are found, what proportion are over 100 years old?

The table below summarises data from news articles collected on Wikipedia about very old bottles with messages that have been found. However, only data on bottles over 25 years old has been collected, presumably because older bottles are more newsworthy.

So, we needed to estimate the number of 0- to 25-year-old bottles with messages ourselves – here’s how we did this.

The table shows that fewer bottles with messages are found as they get older. Messages in bottles degrade over time, which means the bottles have an increased chance of breaking and sinking, or just getting covered in layers of sediment. Plotting this data in the graph below helped us see the trend in the ages of found bottles more clearly.

We drew a line to match this observed trend in the ages of found bottles. This red line in the graph corresponds to the equation:

This equation provides an estimate of how many bottles have been found for any specific age range (where 25 = 0-to-25, 50 = 25-to-50 and so on). We are interested in the the 0- to 25-year-old bottles, so the equation suggests 46 bottles have been found in this range.

Adding up this and all of the numbers in the table gives a total of 106 bottles found, of which 12 are over 100 years old, and 12/106 is about one in ten.

Recapping the above, we have that: (A) one in ten bottles with messages are found, of which (B) one in ten are over 100 years old. Bringing these results together using the multiplication rule, we estimate the chance of a message in a bottle being found and it being over 100 years old to be (1/10)×(1/10) = 1/100.

So, if there are 100,000 bottles with messages floating around the oceans waiting to be found, we’d expect 1,000 of these to be found and be 100 or more years old. Assuming anybody in the world is equally likely to find one of these, with 8 billion people currently, that’s about a one in 8 million chance of you finding one – pretty unlikely.

However, some people are more persistent at message-in-a-bottle hunting than others. Following the paths of ocean currents (known as gyres) could provide clues on where to look.

Specifically, peninsulas or islands intersecting with these gyres could be good spots. For this reason, it has been suggested the Caribbean islands are ideally placed for finding bottles as they lie on the path of the North Atlantic Gyre. Which seems like a great reason to travel to the Carribean!

But let’s also spare a thought for the poor soul stranded on their desert island, who surely won’t appreciate the low odds of their SOS being found.

 

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

The post What is the chance of a message in a bottle being found? appeared first on Mapping Ignorance.

Sewers’ overlooked role in warming

The team estimates that sewers worldwide emit 1.18 to 1.95 million tons of methane annually. This reveals that global wastewater management has a far more significant impact on greenhouse gas emissions and global warming and is crucial for improving greenhouse gas emissions accounting and promoting global emissions reduction.

Challenging assumptions about sewer emissions

It is generally assumed that the limited residence time of wastewater in sewer networks is not conducive to methane formation, and that the overall emissions are difficult to monitor and quantify. Therefore, in current greenhouse gas inventories by the IPCC and various countries, urban sewers are considered a negligible source of methane emissions, assumed to be zero.

However, sewage is rich in biodegradable organic matter, and anaerobic conditions are prevalent in sewer networks, providing a viable environment for methane formation.

From SeweX model to global tool

Professor Yuan’s research team has been dedicated to developing innovative solutions for wastewater systems and environmental biotechnology. As early as 2008, the team successfully developed the SeweX model, which simulates in-sewer physical, chemical and biological processes, including predicting the generation of hydrogen sulfide and methane.

To address the lack of field data for calibrating the methane prediction components of the SeweX model, the team collected data from sewer networks in Australia using a customized online sensor for model calibration and validation.

With the calibrated SeweX model, the team simulated nearly 3,000 different pipeline scenarios with varying structures and operating conditions. Their findings ultimately confirmed that methane generation in sewers is closely related to the wetted pipe surface. They proposed a simplified methane estimation model for sewer systems. With data such as pipe size, slope, designed and actual average dry weather flows, and wastewater temperature, the model can estimate methane emissions from sewers.

The team validated the model using real-world data from 21 cities in Australia, the United States, China and Belgium, thus successfully developing this comprehensive sewer methane emission estimation tool.

Global impact and climate implications

Using this innovative tool, the team estimates that global sewer systems emit approximately 1.18 to 1.95 million tons of methane annually, adding 1.7% to 3.3% to the currently estimated global methane emissions by the waste sector, and approximately 16% to 38% to the estimated overall carbon footprint of wastewater management.

Professor Yuan remarked, “Our research confirms that sewers are not a zero-emission source; rather, they represent a quantifiable source of methane emissions with significant global climate implications. As urban sewers continue to expand, their potential methane emissions will also increase. Therefore, including them in the greenhouse gas accounting system will help improve national greenhouse gas inventories and provide a new entry point for emissions reduction, further advancing the global goal of sustainable development.”

The post Sewers have been hiding a climate problem in plain sight appeared first on Mapping Ignorance.

Here’s the thing about modern computers: at their most basic level, they’re just machines that do really simple things really, really fast. “Really simple things” meaning arithmetic. Things like addition or subtraction —that’s it. A computer processor takes numbers, performs mathematical operations on them according to a set of rules (the program or algorithm), and spits out other numbers.

But here’s where it gets interesting. Those input numbers can represent anything. They can represent the pixels on your screen, the sound waves coming through your microphone, the temperature in your room, the position of a robotic arm, the sequence of a DNA strand. And those output numbers can be translated into anything too—they can become images, sounds, physical movements, answers to questions, predictions about the future.

So when you tap on your smartphone screen, what’s happening? Your tap gets converted into numbers. Those numbers get processed by algorithms—sets of mathematical rules. The resulting numbers get converted into actions: maybe sending a message to a friend, maybe opening an app, maybe adjusting your volume. And it all happens so fast that it feels instantaneous.

This is why calling computers “intelligent machines” isn’t really a metaphor. It’s a pretty literal description. Intelligence, in its most basic sense, is the ability to draw consequences from information. A sunflower turning toward the sun is drawing a consequence (light means grow this way). A squirrel remembering where it buried nuts is drawing a consequence (this spot means food). A person solving a math problem is drawing consequences. And a computer calculating your restaurant recommendation based on your past orders? Also drawing consequences. The difference is speed and scale, not kind.

We hear all the time that technology is progressing fast. But until you look at the actual numbers, it’s hard to grasp just how insane this progress has been. Let’s put it in perspective. In the last two hundred years, maximum travel speed has increased by about a hundred times. That’s impressive—from horse and carriage to high-speed trains, airplanes and spaceships. Human life expectancy has increased by something like two to three times. Average income in developed countries has increased by ten to twenty times. The amount of cargo a single ship can carry is about a thousand times larger than in Napoleon’s time. A farmer can plow about fifty times more land in a week than two centuries ago.

All of these are remarkable achievements, but they absolutely pale in comparison to what’s happened with computers. In the last fifty years alone, the processing power of computers—measured in FLOPS, or floating point operations per second—has increased by a factor of about one million. The most powerful computer today can do in one second what the most powerful computer of the early 1970s would have taken nearly two weeks to do. If similar progress had happened in other areas, we’d be expecting to live about 30 million years, we’d travel from Madrid to Barcelona in less than a second, and we’d all be as rich as Amancio Ortega, the founder of Zara.

One of the most prominent and recent examples are large language models. LLMs are trained on billions of examples of human language—books, articles, websites, conversations—and they learn to predict what word comes next in any given sequence. That sounds simple, but something remarkable emerges from it. By learning to predict words, these models also learn to understand context, to follow instructions, to answer questions, to write poetry, to explain complex ideas, to translate between languages, and yes, to have conversations about technology (are you sure this text has not been created by an AI?). Just a few years ago, if you’d told someone you could have a back-and-forth conversation with a machine that understood what you were asking and responded intelligently and with the nuances of a proficient natural language speaker, they’d have thought you were talking about science fiction.

Here’s something interesting about all this progress: most of it is designed to be invisible. Your smartphone is intentionally built to hide the insane complexity of what’s happening inside it. When you tap an icon, you don’t see the millions of calculations required to open an app. When you send a message, you don’t witness the journey of your words as they get chopped into packets, routed through servers, reassembled, and delivered. And when you chat with an AI, you don’t see the billions of parameters, the layers of neural networks, the massive training runs on supercomputers that made this conversation possible. It just feels natural—like talking to another person, or at least something vaguely person-shaped.

This is actually a lot like what happens in our own brains. You’re not consciously aware of the billions of neurons firing, the electrochemical signals racing along your nerves, the complex processing required to turn light patterns on your retina into the recognition of a friend’s face. You just see your friend and smile. The complexity is hidden so you can focus on what matters.

It’s tempting to look at the last fifty years and try to project the next fifty. If processing power keeps increasing at anything like the current rate, by the end of this century computers will be billions of times more powerful than they are today. The amount of data available for processing will be similarly astronomical. Even if the speed of increase slows by some factor, we probably cannot imagine what will come next. And I mean that literally—not as a figure of speech. The technologies of 2100 will likely be conceived, designed, and built by artificial intelligences that are themselves far more capable than any human engineer. We’ll be in the position of a chess novice watching Stockfish play: we can see that something brilliant is happening, but we can’t fully understand how or why. Think about it. Stockfish doesn’t just beat humans at chess—it plays in ways that humans find baffling. It makes moves that look like mistakes but turn out to be genius. It exploits patterns too subtle for any human to notice. Now imagine that kind of superiority applied to every field of human endeavor (or most of them): engineering, medicine, science, art. The results will be things we literally cannot conceive of today.

Remember Heidegger and Ortega from our first article? Heidegger worried that technology reduces everything to resources, including us. Ortega saw technology as the expression of our human project of self-invention. These new inferential prostheses—these computers and AI systems—bring that debate into sharp focus. Are we becoming resources for our own machines, feeding them data so they can get smarter and more useful? Or are we extending ourselves in ways that Ortega would celebrate, using these tools to become more fully human? The answer, as with most philosophical questions, is probably both. We are both shepherd and cyborg, both threatened and enhanced by our own creations.

What’s certain is that the inferential prostheses we’ve built are now so powerful that they’re starting to draw consequences we couldn’t draw ourselves. They’re finding patterns we couldn’t see, making predictions we couldn’t make, solving problems we couldn’t solve. And they’re doing it at speeds that make our own cognition look like molasses in January.

This doesn’t mean we’re obsolete. It means our relationship with our tools is changing. Language, our first prosthesis, didn’t make us less human—it made us more human. It opened up new worlds of thought, communication, and culture. These new prostheses will do the same, but in ways we’re only beginning to understand. The question isn’t whether to embrace them or reject them. The question is how to use them wisely—how to stay awake to their dangers while remaining open to their possibilities. Heidegger would have us step back and let things be. Ortega would have us step forward and keep inventing ourselves. After all, they’re our prostheses. They’re extensions of us. The future they’re building is, for better or worse, our future too.

The post A mini-philosophy of technology (3): Computers as our ultimate inferential prostheses appeared first on Mapping Ignorance.