Clouds stream on distant exoplanet

Astronomers have used the James Webb Space Telescope to observe a far-off planet in stunning detail, just by studying the tiny fraction of the light from its star that is filtered through the planet’s atmosphere. The team found that thick clouds continuously form on the night side of WASP-94 A b, which is 210 parsecs (690 light years) away from Earth. As winds carry the clouds from the cool night side to the lit day side, they quickly dissipate. The results show how tiny differences in starlight’s spectrum between the start, middle and end of a planet’s transit in front of its star can improve researchers' understanding of planetary atmospheres.

The brain’s code seems to be in constant flux. Neuroscientists are baffled

A long-held dogma of neuroscience states that certain brain cells respond in the same way to the same thing. But that theory’s dominance is being called into question. Researchers have reported evidence for neurons in mice changing how they respond to stimuli over time, a phenomenon dubbed ‘representational drift’. But why drift occurs, and what purpose it might serve, is still unclear. And debates swirl around other big questions, such as: how is the brain able to generate stable behaviours if the activity of its neurons keeps changing?

Blogger Comments:

I — The Myth of the Stable Neuron

Why neuroscience cannot decide what its basic units are

Neuroscience still quietly inherits a metaphysics in which the brain is composed of stable units—neurons—that reliably “do things.” This assumption is rarely defended; it is simply embedded in experimental design, analysis pipelines, and interpretation.

The logic is familiar:

• a stimulus is presented
• a neuron responds
• repeated trials reveal a stable tuning curve
• therefore, the neuron “encodes” something

This produces a deceptively clean picture: a world of features mirrored by a world of neural tokens.

But this picture is not discovered—it is constructed by the experimental frame itself. It depends on:

• discretising continuous activity into unit responses
• holding behavioural context fixed
• assuming that temporal variability is noise around a stable mapping

The problem is not that this model is wrong in detail. The problem is that it defines correctness in advance of what counts as variation.

Once long-term recordings become possible, the supposed stability begins to dissolve—not dramatically, but systematically. The neuron stops behaving like a unit of function and starts behaving like a participant in a shifting coordination field.

At that point, neuroscience does not revise its ontology. It renames the anomaly.

Which brings us to “representational drift.”

---

II — Representational Drift is Not an Explanation

It is a label for the failure of a model that insists on stability

“Representational drift” is often treated as a discovery: neurons change their tuning over time even when behaviour is held constant.

But notice what must already be assumed for this to be a “problem”:

• that there exists a stable representational mapping
• that neural populations are meant to preserve it
• that time should not affect identity of encoding

Drift is then defined as:

deviation from a presumed invariant code

But this is not an explanation. It is a diagnostic term for when a representational model stops fitting its own output.

The deeper issue is that “representation” does not degrade gracefully under empirical stress. It either holds, or it becomes a residual category for everything that no longer behaves as expected.

So the field responds by multiplying distinctions:

• drift
• noise
• plasticity
• remapping
• context dependence

Each term is an attempt to protect the original assumption:

that there must still be a stable code somewhere underneath.

But what if that assumption is optional?

What if the system never contained a stable code in the first place?

---

III — Edelman Already Solved the Problem (And Was Ignored Anyway)

Population dynamics without the need for representational units

Neural Darwinism (Neuronal Group Selection Theory) already displaced the unit-based picture decades ago.

Its core move is simple but radical:

• neural function is not assigned to fixed units
• it emerges from selection over dynamic populations
• stability is a property of recurrent selectional outcomes, not enduring elements

In this frame:

• variability is not a deviation from function
• it is the medium through which function stabilises

Which makes a crucial difference:

what mainstream neuroscience calls “drift” is not a problem in Edelman’s ontology, because there is no requirement for fixed representational identity to begin with.

Yet instead of adopting this shift, the field largely absorbed population language while retaining representational expectations.

This produces a hybrid structure:

• population-level measurements
• representational-level interpretation
• unit-level intuition

And hybrids of incompatible ontologies do not produce synthesis—they produce recurring paradox.

“Drift” is one such paradox.

---

IV — The Stability Problem is Misframed

Why the question “how does behaviour remain stable?” assumes its own answer

The central puzzle in drift literature is:

If neural representations change, how does behaviour remain stable?

This question appears deep, but it quietly assumes:

• behaviour is stable in a way that requires neural invariance beneath it
• neural activity is responsible for preserving that stability

But this reverses the dependency.

Behavioural stability is not something the brain “maintains” against variability. It is an emergent regularity of:

• repeated coupling
• constrained environments
• embodied action
• and re-instantiated neural dynamics

Stability is not stored. It is enacted repeatedly under changing internal configurations.

The confusion arises because neuroscience keeps looking for:

a stable object inside the system that guarantees stability outside it

But there is no such object.

---

V — The Real Crisis: Neuroscience Does Not Know Its Ontology

Why “drift” keeps returning no matter how often it is explained

The persistence of “representational drift” as a concept is not evidence of mystery in the brain. It is evidence of instability in the explanatory framework.

Neuroscience currently oscillates between three incompatible commitments:

1. Neurons as stable encoders (classical tuning curve model)
2. Populations as dynamic systems (Edelman-compatible structure)
3. Representation as explanatory currency (still dominant language)

These cannot all be true at once.

So when data violates (1), the field shifts to (2) but interprets it through (3). When (2) resists (3), the mismatch is labelled (drift, noise, remapping, context dependence).

Nothing ever resolves because the underlying issue is not empirical—it is ontological.

The system is attempting to explain:

dynamic, continuously re-instantiated coordination fields

using:

a vocabulary built for stable symbol-like mappings

That mismatch is the real drift.

---

Coda — What Would Have to Change

A coherent neuroscience would not necessarily need a new discovery. It would require a refusal:

• refusal of neurons as privileged carriers of function
• refusal of representation as default explanation
• refusal of stability as hidden substrate rather than emergent regularity

This is not a call for vagueness or mysticism. It is a call for consistency with the very data neuroscience already produces.

The irony is that the field is not lacking evidence.

It is lacking permission to reinterpret its own evidence without preserving its inherited metaphysics.

India’s DNA map uncovers millions of missing genetic variants

India’s biggest gene-sequencing effort has shed new light on the diversity of the population, identifying nearly 130 million genetic variants, almost a third of which have not been reported previously. The GenomeIndia project analysed the whole genomes of almost 10,000 people, uncovering 44 million variants that weren’t already in global scientific databases. The study also revealed genetic risk factors in some populations, such as variants in genes that affect how the body processes certain drugs, variants linked to anaesthesia-related complications and extremely high levels of genetic homozygosity — when individuals inherit identical forms of a gene at a particular chromosome location from both parents. This can be a risk factor for recessive genetic diseases.

Did Homo erectus and Denisovans mate? Tooth proteins hint at ancient trysts

Ancient proteins extracted from the teeth of Homo erectus individuals that lived in China suggest that the group might have interbred with Denisovans, another archaic human species. Researchers used enamel proteins from six H. erectus individuals and identified an amino-acid variant that’s previously been seen in Denisovans. The group also identified two amino-acid sequence variants shared by all six individuals that set H. erectus apart from humans and other human relatives — something that has proved hard to find in the past.

Ice core reveals longest-ever continuous record of Earth’s climate

A 2.8-kilometre-deep ice core has yielded the longest continuous record of Earth’s climate and atmospheric conditions, stretching back 1.2 million years. Data from the core show how the concentration of carbon dioxide in the atmosphere tracked changes in global temperatures across multiple cycles of climate change. The core covers a period in Earth’s history during which ice ages became less frequent but more brutal. Researchers are hoping the oxygen isotopes and carbon dioxide in the core can provide some hints as to what caused the switch.

Revealed: the mysterious ‘dark’ proteins that might play a big role in biology

‘Dark’ portions of the genome, which weren’t thought to hold instructions for making working proteins, have mostly been excluded from research. Now thousands of the proteins encoded by these genes have an official, new name — peptideins — that marks their inclusion in major gene and protein databases used by the life-sciences community. Dark proteins tend to be very short in amino acid length and lack evolutionary relatives in other organisms, which is part of the reason they have been omitted.

The exotic particles that could finally break the Standard Model

Physicists know that their elegant theoretical description of forces and particles — the Standard Model — must be incomplete, because there are a host of phenomena it cannot explain. But observations continue to confirm the model’s accuracy with ever greater precision. Thank goodness an analysis from an experiment at the Large Hadron Collider at CERN, Europe’s particle physics laboratory, suggests at least one way out: the decay of particles called B meson. A new analysis found that the angle at which the final products emerge from the decay disagree with those predicted by the Standard Model. Evidence for this anomaly has been growing since 2015.

Roman Empire’s collapse created a genetic melting pot in Europe

An analysis of ancient genomes from hundreds of burials in southern Germany reveal that the transition from Late Antiquity to the Early Middle Ages was more complex than the stereotype of northern ‘barbarians’ running roughshod over a Roman Empire in decline. Instead, the genomes point to gradual genetic and cultural shifts that occurred through small-scale migration and intermarriage. Shared patterns of DNA show that northern Europeans “haven’t arrived as mass invaders or hordes or big clans — these are individual families who are already four or five generations on Roman territory”, says population geneticist and anthropologist Joachim Burger, who co-authored the analysis. They probably saw themselves as Romans, he adds.

Do octopus brains work like humans’ — or is there another way to be smart?

Like us and our big-brained vertebrate relations, cuttlefish, squid and octopuses have excellent memories, use tools and are adept problem-solvers; they have a concept of time and are capable of delayed gratification. But they are not like us. “It’s hard to convey how different [the cephalopod brain] is,” says neuroscientist Cristopher Niell. “We just have no idea of how it functions.” Figuring it out could help us to understand where intelligence comes from. But working with a soft, flexible and wily creature — for which we have few ways to relieve pain — is a challenge for neuroscientists.

Olfactory maps overturn our idea of smell

The most detailed map ever of the olfactory receptors in the mouse nose transforms our understanding of the sense of smell. Instead of a handful of broad zones, in which receptors are essentially random, “each receptor adopts a particular position in the nose”, says neurobiologist and study co-author Sandeep Robert Datta. And the arrangement of receptors in the nose is mirrored in the part of the brain responsible for smell. “This means that the maps in the nose and the brain are not two separate problems the system has to solve, but two readouts of the same developmental logic,” says psychologist and experimental neuroscientist Johan Lundström. “This is a landmark paper that overturns one of the foundational textbook models of olfactory organisation.”

Mitochondria can spawn new ‘organelles’ — hinting at how modern cells evolved

When a parasite invades a cell, mitochondria — the cell’s energy producers — respond by shedding their outer layer to form new cellular compartments that digest molecular trash. These new organelles help the parasite to proliferate, possibly because the invader can feed off the degraded material inside the tiny compartments, says immunologist and study co-author Lena Pernas. The findings lend credence to the hypothesis that ancient mitochondria gave rise to the first organelles in complex cells by shedding their outer layer to form new membrane sacs.

Did kraken-like octopuses rule Cretaceous seas? Massive jaw fossils offer clues

In the age of dinosaurs, giant octopuses — sometimes named krakens after the mythological monsters — might have grown to nearly 19 metres in length. The estimate is based on fossilised jaws, which researchers say show patterns of wear that came from devouring animals that had hard shells and skeletons. The team suggest that these giant cephalopods might have sat at the top of the marine food chain in the Cretaceous period alongside huge marine reptiles such as mosasaurs. But other researchers say that estimates about the size of the soft-bodied creatures, and how they hunted, should be taken with a pinch of salt.

Newfound brain network is a ‘secret system’ made of helper cells

Brain cells called astrocytes form extensive networks in the mouse brain that are similar in some ways to those formed by neurons. Researchers have compiled a whole-brain 3D map of astrocyte networks — the first of its kind, they say. The atlas reveals how webs of these cells connect far-flung regions of the brain, allowing them to exchange molecules with each other over long distances.

A Powerful New ‘QR Code’ Untangles Math’s Knottiest Knots

Mathematicians have long tried to untangle exactly what makes a knot a knot. There are several ways to define a knot’s characteristics, referred to as ‘invariants’, but “most invariants are either very strong but impossible to compute, or easy to compute but very weak,” says mathematician Daniel Tubbenhauer. Now researchers Dror Bar-Natan and Roland van der Veen have proposed what they dub “a fast, strong, topologically meaningful, and fun knot invariant” — or Θ for short — that can describe even very complex knots as a rather lovely 2-dimensional pattern.

Sperm whales’ communication closely parallels human language, study finds

The structure of sperm whales’ communications has close parallels with the phonetics of some human languages. The whales (Physeter macro­cephalus) communicate using a series of clicks called codas. The animals can differentiate the sound by changing the click’s length or using rising and falling tones, which researchers found follow patterns that resemble those used in human languages such as Mandarin and Slovenian. “We’re starting to see that these signals are organized in ways we didn’t fully appreciate before,” says behavioural ecologist Mauricio Cantor.

Human evolution sped up after farming

An analysis of DNA evidence from more than 15,000 ancient humans has revealed that human evolution has accelerated over the past 10,000 years. Researchers identified almost 500 gene variants that evolved through natural selection in ancient European and Middle-Eastern people after the dawn of agriculture. Many of those variants are linked to the resistance to diseases, such as tuberculosis. Accelerated evolution could reflect the intensification of lifestyle changes that started in the Neolithic period, such as new foods and pathogens, says population geneticist David Reich.

Sunbirds suck nectar with supple tongues

Sunbirds (Nectariniidae) drink nectar from flowers by generating suction with their tongues — the first example of a vertebrate creating suction without changing the shape of its mouth. As they drink, the birds repeatedly press their tongue against the roof of their beak and then release it. This motion flattens and re-forms a v-shaped groove along the length of the tongue, which researchers believe creates suction that pulls nectar into the birds’ mouths.

Quantum simulations verified by experiments for the first time

For the first time, physicists have matched detailed quantum-computer simulations to experimental data gathered from work with solid materials. Two teams of physicists achieved the feat independently: one simulated the physical properties of a magnetic material, such as its heat capacity, and the other modelled a different material’s response to being excited into a range of energy states. Both agreed with experimental data.

Genetic record for domestic dogs pushed back by 5,000 years

Researchers have identified the earliest known dog genomes, which push the genetic record for dogs back by more than 5,000 years. They recovered these genomes from remains of between 14,000 and 16,000 years old found at archaeological sites that span Europe and the Middle East. The team also identified an early domestic dog population (Canis lupus familiaris) that spanned Western Eurasia and was kept by diverse human hunter-gatherer groups. The findings show that dogs were exported and exchanged by various human groups, underlying dogs’ importance to early communities with different ways of living.

‘Nuclear clocks’ tick closer to reality

After decades of research, physicists are confident that the first tick of a ‘nuclear clock’ is imminent. This device would keep time by measuring energy transitions in the radioactive isotope thorium-229, which could make it the most precise clock on the planet. Last week, scientists at the American Physical Society Global Physics Summit compared notes on their progress toward assembling the clock’s components, including efforts to build a laser that can make the clock ‘tick’.

Debate explodes over age of key South American archæological site

An archæological site that overturned the history of humans in South America might just do it again. Radiocarbon dating pegged artefacts from Monte Verde, in southern Chile, at 14,500 years old — suggesting that people arrived along the continent’s coast, before the ‘Clovis people’ travelled through an ice-free corridor in North America. Not so, says a new study: those artefacts were mis-dated and originated no more than 8,200 years ago.

New psychedelic fungus rewrites origins of magic mushrooms

A newly described species of ‘magic mushroom’ could upend a popular theory of when psychedelic fungi popped up around the world. Researchers had thought that the magic mushrooms that grow in southern Africa were Psilocybe cubensis, the same species that grows in the Americas. However, closer inspection revealed that the African mushrooms are a separate species, now named Psilocybe ochraceocentrata, and last shared a common ancestor with P. cubensis approximately 1.5 million years ago. These findings scupper the hypothesis that P. cubensis was inadvertently introduced to the Americas by 16th-century settlers, but the research offers no clues as to its origins across the Atlantic.

‘Virtual cell’ captures most-basic process of life: bacterial division

Researchers have created a 3D simulation that models DNA replication, cell division and nearly every chemical reaction in a living bacterial cell. This ‘virtual cell’ isn’t a totally faithful recreation of the organism — the team used placeholders for some genes with unknown functions, for example. But it could help researchers understand how the mix of molecules in a cell gives rise to actual life, says computational biophysicist and study co-author Zane Thornburg.

First ‘half Möbius’ carbon chain could unlock strange physical properties

Chemists have synthesised a new type of carbon-based molecule with an unprecedented twist in its structure. The team calls the looped molecule a ‘half-Möbius’, inspired by the Möbius strip — a twisted loop with one continuous surface. In the half-Möbius molecule, the chain of atoms is twisted by 90° to make the loop, instead of the full 180º seen in a standard Möbius strip. The molecule can exist in two versions depending on whether it twists left or right. The versions differ in chirality, meaning that they are mirror images of each other.

Galileo’s handwritten notes found in ancient astronomy text

Around 1590, Galileo Galilei filled the margins of a copy of Ptolemy’s The Almagest with his musings — and they have just been rediscovered. Historian Ivan Malara recognized the storied astronomer’s handwriting in the book, which was held in a library in Florence. The contents, including a transcribed psalm, could shed light on the intellectual origins of Galileo’s groundbreaking view of the Solar System.

Earth’s oldest crystals suggest an early start for plate tectonics

Chemical evidence from ancient crystals suggests that Earth’s tectonic plates had already started moving 3.3 billion years ago, relatively early in our planet’s history. By analysing zircon crystals recovered from Western Australia — the oldest-known fragments of Earth rocks — researchers also found that the ancient Earth could have contained more oxygen, and possibly more water, than suspected. The movement of tectonic plates and higher-than-expected oxygen levels suggest that conditions on Earth could have been more conducive to life during this period than previously thought.

Why ‘quantum proteins’ could be the next big thing in biology

Researchers have found ways to turn fluorescent-protein labels — coloured tags used to track the inner workings of a cell — into qubits, the basic units of quantum information. In this form, they can be used as quantum sensors, which could reveal cellular activity and detect molecules associated with disease with unprecedented detail. The development of protein quantum sensors is at an early stage, but could progress quickly, experts say: the approach has been shown to work in principle and the necessary equipment is standard fare.

Neanderthal dad, Sapiens mum: study reveals ancient procreation pattern

Genetic evidence from three Neanderthal (Homo neanderthalensis) specimens suggests female Homo sapiens and male Neanderthals mated more often than did male H. sapiens and female Neanderthals — an imbalance that helped to shape the human genome. Researchers found that the X chromosomes from the specimens had, on average, 62% more sapiens DNA than non-sex chromosomes, which could be explained by a bias in mating. The bias could have been down to mate availability or cultural sanctions for certain combinations, says population geneticist Sohini Ramachandran.

Blogger Comments:

Here's a hint for these geneticists from anthropology: In tribal warfare, the victors kill the men and absorb the women into their tribe.

40,000-year-old German artefacts may display written language precursor

A series of notches carved on a piece of mammoth ivory roughly 40,000 years ago could be an early ancestor of cuneiform, one of the oldest-known forms of written language. “These sign sequences go beyond decoration that was aesthetically pleasing to particular individuals,” says linguist Christian Bentz. He co-authored a new study of the object and more than 200 other Stone Age artefacts that bear these signs. The study found statistical patterns in the Stone-Age marks that are similar to early proto-cuneform — but quite different to the information-dense writing systems of today.

Ancient rocks point to an early start for the Great Unconformity

Fresh evidence from rocks in China is contributing to debate about what caused the ‘Great Unconformity’ — a one-billion-year gap in the geological record that exists all over the world. Theories abound: it might have been caused by abrasive worldwide glaciers during a period of ‘Snowball Earth’, or by erosion after the break-up of the supercontinent Rodinia. Tantalisingly, the period precedes a sudden proliferation of life on Earth, known as the Cambrian explosion. But new evidence suggests that most of the erosion that caused the gap happened long before all of that.

Glass tile can store data for millennia

Researchers at Microsoft have created a data-storage system that can remain readable for at least 10,000 years — and probably much longer. The team used a high-energy laser to imprint deformations into a 3D chunk of borosilicate glass, the kind used in ovenware. Each deformation encodes data that can be read out using a microscope. A 12-centimetre wide, 2-millimetre-thick square of the glass can securely store nearly 5 terabytes of data, the equivalent of around 2 million printed books.

An ultrashort laser pulse creates tiny modifications to the optical properties of the glass called voxels, which encode data. A square of glass containing 301 layers of voxels had a storage capacity of 4.8 terabytes.

Giant virus hijacks cells’ protein-making machinery to multiply wildly

Scientists have identified a giant virus that can hijack a host cell’s protein-making machinery to churn out copies of itself — the first experimental evidence that viruses can co-opt this particular system, which is typically associated with cellular life. To take control, the virus attaches a three-protein complex to the host’s ribosomes — part of the apparatus cells use to make proteins — which gives viral RNA preferential access. Researchers suggest that the virus makes this protein complex using genes that it ‘stole’ from hosts early in its evolutionary history.

Canadian fossil reveals one of the first plant-eating animals

A fossilised skull dated to more than 300 million years ago has been identified as belonging to Tyrannoroter heberti, a newly described ‘microsaur’ species and one of the oldest-known vertebrate herbivores. From a 3D reconstruction of the skull, researchers found that the species had specialised teeth that resemble those of other plant-eating animals and that had been worn down from a grinding motion consistent with herbivory. The discovery suggests that vertebrate herbivores emerged around several million years earlier than scientists had previously estimated.

Hunter-gatherers took refuge in European 'water world' for millennia

Ancient genomic data has revealed that inhabitants of the Rhine-Meuse river delta — wetland and coastal areas of modern-day Netherlands, Belgium and western Germany — maintained high levels of hunter-gatherer genetic ancestry for thousands of years after successive migrations from the east transformed most of Europe into farming and animal-herding communities. This group’s eventual mix with communities of people with ancestry from the central Eurasian steppe catalysed the expansion of Bell Beaker culture, which was accompanied by major shifts in the genetic make-up of both Britain and the Rhine–Meuse delta.

Rules of mysterious ancient Roman board game decoded by AI

Researchers say they have figured out the rules of an ancient Roman board game. The team simulated thousands of turns based on more than 100 games that could be played on the linked lines inscribed on a stone found in the Roman settlement of Coriovallum, now Heerlen in the Netherlands. One set of rules best matched the wear-and-tear on the stone: a blocking game they’ve nicknamed Ludus Coriovalli (‘the game from Coriovallum’).

Sibling Rivalry in Baboons

Young chacma baboons (Papio ursinus) vie for their mother’s attention more often when she’s grooming one of their siblings than when she’s free, which suggests that they get jealous of one another. Researchers found that the monkeys were more likely to interrupt their mother if she was grooming a younger sibling or one of the same sex. Their tactics included tantrums, attempts to physically come between their mother and sibling, and even tricks to lure their sibling away from their mother and take their place.

Does the galaxy have dark matter at heart?

The black hole at the centre of our galaxy might actually be a huge blob of dark matter, suggest astrophysicists. A super-dense core of subatomic particles called fermions, spreading out into a vast, diffuse cloud, would match up with observations (and a direct image captured in 2022) equally as well as a black hole, they say. If the Milky Way really is permeated by a cloud of fermionic dark matter — and we’d need better observations to confirm it for sure — it could help to answer outstanding questions about how the whole galaxy behaves.

Does AI already have human-level intelligence? The evidence is clear

When will artificial intelligence systems achieve artificial general intelligence (AGI), when defined as the ability to convincingly imitate the “broad, flexible cognitive competence” of a person? We’re already there, argue four scholars in philosophy, machine learning, linguistics and cognitive science who look back to a seminal 1950 paper by Alan Turing as their guide. Large language models can already chat convincingly, write passable poetry and prose, solve mathematics problems, propose scientific experiments and assist in writing computer code. “The machines Turing envisioned 75 years ago have finally arrived, in a form both more alien and more human than anyone imagined,” write the authors, who argue that recognising this matters for understanding our minds and world.

Blogger Comments:

AGI, Stochastic Parrots, and the Culture That Defines Intelligence

"Furthermore, there is no guarantee that human intelligence is not itself a sophisticated version of a stochastic parrot."

That sentence, from a recent Nature commentary arguing that artificial general intelligence (AGI) may already be here, does more than provoke. It is a hinge — a small linguistic pivot around which a vast conceptual shift quietly turns. To read the claim at face value is to miss the larger, subtler work being done: a redefinition of intelligence itself, human and artificial alike.

---

AGI Already Here? The Nature Argument

The authors claim that large language models and related systems already demonstrate the kind of broad, flexible cognitive competence that Alan Turing imagined in 1950. These systems can chat convincingly, generate prose and poetry, solve mathematical problems, propose scientific experiments, and even assist in writing code. By Turing’s criterion — the imitation game — these capabilities are presented as evidence that AGI is not a distant horizon, but a present reality.

At first glance, this feels startlingly plausible. Chatbots can answer questions with fluency, propose solutions with apparent insight, and mimic reasoning across domains. Yet the claim rests on a subtle, often unspoken manoeuvre: intelligence is defined by performance on tasks we, historically and culturally, consider meaningful. Benchmarks and success criteria are not neutral measures; they are socially stabilised definitions.

---

The Meta Problem: What Counts as Human Intelligence?

The Nature commentary is compelling because it leverages unexamined assumptions about human intelligence. Intelligence is treated as stable, measurable, and largely symbolic: the ability to communicate, reason, and solve problems in literate, analytic ways. But this proxy omits much of what humans actually do: navigate risk, act within moral or normative frameworks, participate in embodied practices, and respond to real-world consequences.

By suggesting that humans might themselves be “sophisticated stochastic parrots,” the article flattens the human into a process of pattern extraction, a subtle but radical deflation that allows machines to be measured on the same plane.

“If humans are sophisticated parrots, why can’t machines be too?”

The deeper meta-move is epistemic: uncertainty about the nature of human intelligence is leveraged to lower the threshold for recognising intelligence in machines. What appears humble is actually a strategic repositioning.

---

The Circularity of AI Culture

Here we encounter a deeper structural point: intelligence, as currently defined in AI discourse, is co-constituted by the culture of its developers. Consider the loop:

1. Developers set tasks — benchmarks, coding challenges, dialogue prompts — based on what they value and can measure.

2. AI systems perform these tasks, optimised to succeed.

3. Success validates the AI as “intelligent.”

4. That validation shapes the culture, reinforcing which tasks matter, which challenges are prioritised, and what counts as intelligence.

In short: the tasks define intelligence, the AI performs the tasks, and the AI’s performance confirms the validity of those tasks.

This is not merely a critique of metrics; it is a structural observation about mutual actualisation. Intelligence is not simply a property of the machine; it emerges from the interaction between human priorities, institutional practices, and technological affordances.

The circularity is invisible because it is internal to professional practice. From the outside, performance looks natural, inevitable, and “objective.” Yet it is profoundly contingent, culturally and historically situated.

---

Correlation, Structure, and the Flattening of Intelligence

The article reinforces this perspective with the line:

"All intelligence, human or artificial, must extract structure from correlational data; the question is how deep the extraction goes."

This is a masterstroke of conceptual framing. Intelligence is reduced to pattern recognition and abstraction. Depth, not kind, becomes the relevant metric. Qualitative, embodied, and normative aspects of human cognition are quietly flattened into a single continuum.

The rhetorical power is subtle but immense. Humans and machines are rendered comparable not because they share experience or consequence, but because they share a capacity for structural extraction. Depth becomes the axis along which competence is measured; the stakes, the embodiment, the meaning, and the lived consequences of action are bracketed away.

“Once the debate is framed around depth of extraction, scale becomes destiny.”

The human is reconstructed to fit the machine, and the machine is praised for mirroring the flattened human. The explanatory direction matters: we do not evaluate the machine against the human; we evaluate both through the lens of pattern extraction, and the human is quietly redefined to fit.

---

Implications and Takeaways

Viewed meta-analytically, the Nature article does less reporting than cultural reconfiguration. Intelligence is not a pre-existing property; it is co-constructed through human practice, task design, and perceptual validation. Declaring AGI “already here” is thus as much a reflection of cultural priorities as it is a statement about technological capacity.

Two consequences follow:

1. The question of AGI shifts

   From:

   “When will AI become intelligent?”

   To:

   “Which aspects of human intelligence do we prioritise, for whom, and under which cultural regimes?”

2. The human is subtly redefined
   By flattening human intelligence into a continuum of depth in pattern extraction, our own conception of mind, agency, and cognition is quietly reshaped. Machines are not just performing tasks — they are participating in a mutual recalibration of intelligence itself.

---

A Closing Reflection

The Nature commentary does not merely announce AGI; it reframes what counts as human intelligence. By normalising structure extraction as the essence of cognition, it creates a space in which machines appear not merely competent, but generically intelligent.

Yet the most compelling intelligence may reside not in machines, but in the capacity to perceive and critique the loop by which intelligence is defined. Recognising the co-constitution of human and machine intelligence — the mutual shaping of definitions, priorities, and validation — may be the most reflexively powerful act of cognition we can perform today.

“Perhaps the most interesting intelligence is not in the machines at all, but in the ways we define, measure, and collectively actualise intelligence through our own cultural practices.”

In the end, the stochastic parrot is not only a mirror for AI, but also a mirror for us. The intelligence that matters is the intelligence that notices the mirror — and steps back, just long enough, to see the loop itself.

430,000-Year-Old Wooden Tools Are the Oldest Ever Found

Wooden objects carrying the marks of carving and use could be the oldest wooden tools ever found. Researchers dated the artefacts, found in what is now Greece, to 430,000 years ago — and suggest they might have been made by early Neanderthals or their ancestors, Homo heidelbergensis. A separate study describes 480,000-old flint-knapping tools made from antler and elephant bone, from what is now the United Kingdom. Organic artefacts are a rare find because they’re less likely to endure than stone, and show how “bone and wood were probably more valuable for our ancient ancestors,” says archaeologist Thomas Terberger. “Imagine how many tools you can make from a single large bone of an elephant.”

First animals known to fake the smell of flowers

Larvae of the European blister beetle (Meloe proscarabaeus) mimic the scent of flowers to find a ride to their next meal. The larvae produce floral-smelling compounds to lure bees, then hitchhike back to the bees’ nest and eat the eggs they find there. This trickery is the first known example of an animal imitating the scent of a flower.

Schrödinger’s cat just got bigger: quantum physicists create largest ever ‘superposition’

In the iconic quantum double-slit experiment, a photon acts like a wave and interferes with itself — illustrating the mind-boggling concept that particles can exist in a ‘superposition’ of possible locations at once. But just how big can such a quantum object be? At least as big as a blob of around 7,000 sodium atoms, it turns out. Using a painstakingly constructed experimental set-up, researchers managed to observe tell-tale interference patterns for sodium clusters some 8 nanometres wide, with superpositions that stretched across 133 nanometres. The clusters are as big as some viruses, raising the tantalising possibility of putting biological matter through the same experimental set-up.

Blogger Comments:

This is an impressive and painstaking experiment. Demonstrating clear interference patterns for clusters of around 7,000 atoms, spatially separated by more than 100 nanometres, represents a real extension of the regimes in which quantum descriptions can be experimentally sustained. As experimental control of isolation, coherence, and interferometric precision, the work deserves genuine admiration.

What is worth handling carefully, however, is how such results are often presented.

When articles speak of objects “existing in a superposition of locations at once”, or frame the experiment as probing whether quantum mechanics “still applies” at larger scales, a subtle shift occurs. Formal features of a successful theoretical description begin to be treated as literal claims about what the system is, rather than about how it can be described under tightly controlled conditions.

From a more structural perspective, a superposition is not an ontological state of affairs. It is a theoretical potential: a space of possible outcomes defined relative to a particular experimental arrangement. The interferometer does not reveal a sodium cluster to be “in many places”; it actualises a phenomenon whose meaning is inseparable from the construal that makes it observable.

Seen this way, the familiar question — “where does the quantum world give way to the classical?” — is slightly misplaced. What changes is not the world itself, but the stability of the conditions under which certain descriptions remain coherent. Quantum mechanics does not abruptly fail at larger scales; rather, it becomes progressively harder to maintain the isolation and precision required for quantum descriptions to remain usable.

The real achievement of experiments like this is therefore not that they show ever-larger objects to be “really” quantum, but that they map how far we can extend a powerful theoretical construal before the practical conditions that sustain it dissolve.