Transcription
There is a small creature sitting on a piece of fruit somewhere near you right now. Its entire life lasts roughly 60 days. It cannot read, cannot plan, cannot worry about the future. It is drawn to the sweet and repelled by the bitter. It courts mates with a song it has never been taught. It sleeps. It remembers, however briefly, where danger lives. And for a very long time, we had no idea how any of that worked. Not really. Not at the level that matters. Not at the level where the question ceases to be philosophical and becomes something you can see.
The brain, any brain, has remained fundamentally illegible for most of human history. We have known for centuries that neurons exist, that they pass signals, that they form networks of unimaginable complexity. We have built scanners that light up regions of activity, surgeries guided by maps of function, drugs that nudge chemical systems we barely understand. But the wiring diagram, the actual complete neuron by neuron, synapse by synapse schematic of how any brain is physically built, that had never been produced. Not for an animal of any meaningful complexity, not until 2024.
In October of that year, a global consortium of researchers published across nine simultaneous papers in the journal Nature, the first complete connectome of an adult brain. Every neuron, every connection, every predicted synapse type, whether activating or suppressing, the organism was a fruit fly, a creature with 139,255 neurons and more than 50 million synaptic connections. And the implications of that map and of what researchers found inside it stretch far beyond the insect sitting on your peach. This is the story of how we arrived at that map. What it took to build it, what it has already revealed about the logic of animal behavior, and what it forces us to confront about the nature of thought itself in creatures as small as a grain of sand and in minds as large as our own.
Mapping the human brain requires a kind of thinking most of us were never formally taught. Not just an understanding of biology, but the ability to look at a system of 100 billion connections, follow the logic inside it, and actually make sense of what you are seeing. That is the kind of thinking I have been building through Brilliant, who are sponsoring this video. What I love about their course on how AI works is that it walks you through exactly how neural networks process and store information, and it does it by putting you inside the problem rather than just explaining it. You learn by solving, not by watching. That difference is what makes ideas actually stick. The courses are built by educators from MIT, Harvard, and Stanford. And they are designed for anyone from age 10 to 110. So whether this kind of thinking is new to you or something you're building on, there is a clear path through. To try everything on Brilliant free for a full 30 days, head to brilliant.org/omni or scan the QR code on screen. Brilliant has also given Omni viewers 20% off an annual premium subscription, which gives you unlimited daily access to everything on the platform. That is brilliant.org/omni.
Now to understand what this map took to build, we have to go back much further than 2024. Long before anyone knew what a neuron was, people understood that the brain was the seat of something. The ancient Egyptians dismissed it. Pharaohs were mummified with their organs carefully preserved, but the brain was scooped out through the nostrils and discarded. The heart was what mattered. The heart was where the person lived. It was Hippocrates in the fifth century B.C.E. who argued that the brain was the organ of sensation and intelligence. He was mostly ignored for centuries, but the idea persisted, slowly gathering weight until the invention of the microscope in the 17th century began to reveal the fine structure of living tissue.
Even then, the neuron remained mysterious. The cell theory that dominated biology through the 19th century assumed that all tissues were continuous, that the nervous system was a single unbroken web like a fishing net through which signals simply flowed. The Spanish anatomist Santiago Ramón y Cajal challenged this directly through meticulous hand-drawn illustrations made in the 1890s. Cajal argued that the nervous system was composed of discrete individual cells, neurons, each separate from its neighbors communicating across tiny gaps rather than through direct physical fusion. He was right. And the gaps he described, synapses, as Charles Sherrington later named them, turned out to be the most important real estate in biology. Cajal won the Nobel Prize in 1906. But even his extraordinary maps, drawn with the patience of a watchmaker and the precision of an artist, were incomplete. They showed shapes and types and general arrangements. They did not show, could not show, the actual wiring, which neuron connected specifically to which, with what strength, and in what direction. A map of the streets without any indication of the traffic, the intersections, or whether the roads were one-way or two.
The field of neuroscience spent the 20th century accumulating tools. The electroencephalograph, the patch clamp technique, which allowed researchers to record the electrical activity of a single neuron. Functional magnetic resonance imaging, which measures blood flow as a proxy for brain activity across entire regions. These tools revealed enormous amounts, but they all operated at a level of resolution that kept the fundamental question just out of reach. How does structure create function? How does the physical wiring of a brain generate behavior? That question required something different. It required a complete circuit diagram.
The first organism to have its nervous system fully mapped was not a fly and not a mouse. It was a worm. *Caenorhabditis elegans* is a transparent roundworm roughly 1 mm in length that lives in soil and has been a model organism in biology since the 1960s. Its appeal is its simplicity. In 1986, Sydney Brenner, John White, and colleagues at the Medical Research Council Laboratory of Molecular Biology in Cambridge published the complete wiring diagram of the *C. elegans* nervous system, 302 neurons connected by approximately 5,000 synapses. The work had taken more than a decade of painstaking manual annotation of electron microscope images. It was a landmark paper, and Brenner later received a share of the Nobel Prize in Physiology or Medicine for his work on *C. elegans*.
Those 302 neurons can be studied in their entirety. Researchers can look at every connection, every circuit, follow a signal from sensory input to motor output. And they have, over the decades since 1986. The *C. elegans* connectome has been used to understand chemosensation, thermosensation, pain responses, feeding circuits, and the neural basis of learning in an animal that, by all outward appearances, does not seem to do much learning. But 302 neurons is a small canvas. The circuits that govern even the simplest worm are tractable. But they don't tell you much about how circuits scale. How the logic of neural computation changes as you add orders of magnitude of complexity. For that, scientists needed a larger brain.
For decades, the leap from 302 neurons to anything resembling a vertebrate brain seemed impossibly far. The electron microscopy required to image synapses at nanometer resolution was extraordinarily slow. Slicing brain tissue into thin sections by hand, imaging each section, then painstakingly reconstructing the three-dimensional structure from thousands of two-dimensional images. It was the kind of work that could occupy a scientist for an entire career and yield only a tiny fragment of a tiny brain.
The fruit fly changed the calculation. *Drosophila melanogaster*, the common fruit fly, had been a laboratory staple for over a century. Thomas Hunt Morgan used fruit flies at Columbia University in the early 1900s to demonstrate that genes resided on chromosomes, work that earned him the Nobel Prize in 1933. By the early 21st century, *Drosophila* had accumulated one of the richest genetic toolboxes in all of biology. Researchers could activate specific individual neurons, silence them, label them with fluorescent markers, and track their activity in living flies with remarkable precision. A complete wiring diagram of the fly brain would give all of those genetic tools a structural context, a blueprint to read alongside the circuitry.
The fly brain is about the size of a poppy seed. It contains roughly 139,000 neurons, a million times fewer than the human brain. But those neurons support a behavioral repertoire that, in its complexity, continues to surprise researchers. Fruit flies navigate. They learn to avoid odors associated with harm. They court potential mates with elaborate species-specific songs produced by vibrating a wing. They sleep, and they are sensitive to caffeine. They can become, in a meaningful physiological sense, intoxicated by alcohol and they show behavioral sensitization to it, the same phenomenon that underlies human tolerance. As Sebastian Seung, professor of computer science and neuroscience at Princeton University and co-leader of the FlyWire Consortium, noted at Neuroscience 2024, the Society for Neuroscience's annual meeting, 75% of the disease-related genes in humans have direct counterparts in the *Drosophila* genome. The creature is not as alien as it appears.
The origin of the FlyWire connectome traces to a decision made in 2013. David Boech and his team at the Howard Hughes Medical Institute's Janelia Research Campus took an adult female fruit fly brain and preserved it in a hardening chemical solution. Over the following years, they used a specialized instrument to slice the solidified, poppy seed-sized brain into exactly 7,500 ultra-thin sections, each one a fraction of a micron thick, and imaged each section using an electron microscope. By 2018, they had produced 21 million separate microscope images of the fly brain, each revealing cellular structures at nanometer resolution. They made the entire data set publicly available. The scale of that data set was staggering. No individual lab, no small team of scientists could realistically trace every neuron through 21 million images by hand. But the data set existed. The raw material was there.
Mala Murthy, then director of the Princeton Neuroscience Institute, and Sebastian Seung saw an opportunity. Seung had spent years developing the concept of crowdsourced connectomics, using games and citizen science to deploy human pattern recognition at scale. His earlier project, EyeWire, had engaged online gamers in tracing neurons through a mouse retina, demonstrating that motivated non-specialists could contribute meaningfully to the work of neural reconstruction. The FlyBrain data set offered a chance to take that model to a new level of ambition.
The project they launched, FlyWire, combined multiple approaches into a single integrated effort. Artificial intelligence algorithms were trained to automatically segment neurons in the microscope images, to identify and label each distinct cell through the full three-dimensional volume of the brain. The AI did the heavy lifting, but it made mistakes. So FlyWire recruited proofreaders, neuroscientists from labs around the world, paid annotators, and citizen scientists who engaged with the data through a game-like interface that turned the painstaking business of tracing neural fibers into something that resembled assembling a three-dimensional puzzle. The scale of the human contribution was itself remarkable. As Amy Sterling, executive director of EyeWire and the crowdsourcing manager of FlyWire, described at Neuroscience 2024, an elite group of citizen scientists proofread 18,000 neurons and annotated, added cell type labels to more than 38,000 neurons in the fly brain. In total, the researchers and contributors to FlyWire collectively contributed 33 person-years of effort to proofreading and annotating the AI's output. Without the AI doing the initial work, Seung estimated the project would have required something on the order of 50,000 person-years, more than the entire recorded history of modern neuroscience condensed into a single project.
The result, published in October 2024 across nine papers in Nature, was a complete wiring diagram of the adult fly brain. Every one of the 139,255 neurons connected by more than 54.5 million synaptic connections. The connectome identified 8,453 distinct cell types, 4,581 of which had never been described before. It covered 78 anatomically distinct brain regions called neuropils, each associated with different sensory, motor, or cognitive functions. And it was the first whole-brain connectome to predict the functional role of every connection, whether each synapse was excitatory, promoting the transmission of a signal, or inhibitory, suppressing it. To appreciate the physical scale of what had been mapped, unraveled end to end, the neural wiring of that single fly brain would stretch more than 490 ft. That is longer than four blue whales aligned nose to tail.
A map is only as useful as the questions you bring to it. Even before the formal publication in October 2024, the FlyWire data set had been publicly available, and researchers had begun to use it. By the time the nine Nature papers appeared, the connectome had already been cited in more than 50 published studies. What those studies found, and what the FlyWire team itself reported in the companion analyses, constituted the beginning of something genuinely new in neuroscience.
The first thing the map revealed was not about any particular circuit but about the overall architecture of the fly brain as a network. Analysis of the connectome's topological structure, reported in one of the companion Nature papers, found that the fly brain displays what network scientists call "rich club" organization. In network theory, a rich club refers to a subset of highly connected nodes that are themselves more densely interconnected with one another than their degree of connection would predict by chance. In social networks, this resembles the phenomenon of well-connected individuals who also happen to know each other well. In road networks, it describes major highway hubs that are directly linked to other major hubs, creating a backbone for efficient long-distance traffic. In the fly brain, roughly 30% of all neurons belong to this rich club structure. These neurons are hubs. They receive signals from many different parts of the brain and send signals to many others. Researchers identified subsets of them that appear to serve as integrators, pooling information from diverse sources, and broadcasters, distributing processed signals widely. The existence of this architecture in an insect brain is notable because the same organizational principle had previously been identified in human brain networks using MRI. Work first published by Martijn van den Hout and Olaf Sporns at Utrecht University Medical Center in the Journal of Neuroscience in 2011. The fact that a fruit fly and a human share this deep organizational feature suggests it is not a coincidence of scale or complexity. It may be something more fundamental. A solution that evolution converged on independently. A principle of efficient neural computation that applies whether you are navigating a kitchen counter or navigating a continent.
Beyond global architecture, the connectome revealed unexpected geography. A region called the subesophageal zone, or SEZ, had been poorly represented in previous partial fly brain maps. The full connectome showed that the SEZ is one of the most connected regions in the entire brain. It receives a large fraction of the signals transmitted into the brain from the outside world and it sends nearly all of the output signals that go to motor neurons, the neurons that control movement. The SEZ is, in other words, not a peripheral relay station. It is a central hub for the integration of sensory experience and the generation of action. Its role had been essentially invisible until the complete map made it legible.
The connectome also allowed researchers to compute what they called a "projectome," a map not of individual synaptic connections but of the broader patterns of projection between brain regions. Which areas talk to which and how strongly. The projectome revealed that most projections in the fly brain stay within one hemisphere, but certain categories of neurons are far more likely to project to the opposite hemisphere than others, suggesting that cross-hemisphere integration is a specialized function rather than a general feature of brain organization.
Perhaps the most immediately practical finding came from a computational model built directly on the connectome by researchers at the University of California, Berkeley. Working with the FlyWire team, they constructed a brain-wide model of neural firing based on the connectome's wiring diagram and tested its predictions against known circuits for feeding and grooming behavior. The model accurately reproduced the known circuits, which was encouraging. Then they pushed it further. They asked the model to predict how information about different tastes flows through the brain. The prediction was surprising. It suggested that the neural circuits processing the taste of sugar and the taste of water overlap substantially, a prediction that seemed counterintuitive because these are functionally distinct stimuli. But when researchers tested the prediction experimentally in live flies, they confirmed it. The model built from the structure of the connectome alone had predicted something real about how the brain processes sensory experience. This is what a complete circuit diagram makes possible. Not just description but prediction. Not just anatomy but function.
There is a parallel that scientists have been drawing since the FlyWire project began, and it is worth sitting with. The complete genome of *Drosophila melanogaster* was published in 2000. It was the product of years of work by the Berkeley Drosophila Genome Project and collaborating institutions, and it represented a landmark in genetics: the first time the complete DNA sequence of a multicellular organism with a nervous system had been read from end to end. In the years that followed, that genome became a foundation for thousands of discoveries. Researchers used it to identify genes involved in circadian rhythms, in development, in memory, in disease susceptibility, in aging. The genome was the parts list. With the parts list in hand, science accelerated.
The connectome is, in an important sense, the genome of the brain. It is the structure that underlies everything. Without knowing which gene does what, you cannot understand how an organism develops, responds to its environment, or malfunctions in disease. And without knowing how neurons are physically wired together, which ones connect to which, in what configuration, with what type of synapse, you cannot understand how circuits generate behavior, how learning changes them, or how their disruption produces neurological disease. Anita Devine, a fruit fly neuroscientist at Emory University, who was not part of the FlyWire project but has used its data extensively, put it bluntly when she spoke to the New York Times in October 2024. Her lab was using the connectome for everything they do. For decades, researchers had been studying taste circuits in *Drosophila* without knowing which neurons in the brain were specifically responsible for processing taste signals. The connectome revealed them essentially immediately. What had been an unknown was suddenly visible. The circuit was there in the map, waiting to be read. Martha Patacharia, associate professor of neuroscience at the University of Arizona, described the shift in terms of information flow. That with the connectome, researchers could finally follow how a signal moves from sensory input through the brain's processing stages to motor output in a way that had never been possible before. This kind of tracing is the fundamental act of neuroscience. The connectome made it tractable at whole-brain scale for the first time in a complex organism.
The comparison to the genome is not just metaphorical. It has a specific technical implication. The genome of *C. elegans* was sequenced in 1998. The genome of *Drosophila* followed in 2000. The human genome, the whole thing, was declared complete in 2003, 3 years later. The progression from worm to fly to human took 5 years. No one expects the progression in connectomics to be that rapid. The human brain is roughly 600,000 times more complex than the fly brain in terms of neuron count, somewhere around 86 billion neurons and more than 100 trillion synaptic connections. But the technology is moving faster than it ever has. And the FlyWire project demonstrated a path.
As Moritz Helmstaedter, a connectomics researcher at the Max Planck Institute for Brain Research, told Science magazine in October 2024, "Within the next decade, we should expect to see tremendous progress in connectome mapping and possibly the first full mammalian brain connectome." The zebrafish, whose brain contains about 100,000 neurons, is already a target. The mouse brain, roughly 70 million neurons, is the next major challenge. And in April 2025, an international team including Princeton's Sebastian Seung and the Allen Institute for Brain Science published what they called the Microns Project, the largest and most detailed connectome ever produced of a mammalian brain, covering 1 mm³ of a mouse's visual cortex. That single cubic millimeter contained more than half a billion synaptic connections. 1 mm³ of mouse brain. The whole mouse brain is roughly 500 mm³. The human brain is over a million. The mathematics of this challenge is not hopeful on a time scale of years, but the trajectory is real.
One of the oldest questions in the study of animal behavior is whether behavior is learned or innate, shaped by experience or written into the structure of the organism at birth. This is a false dichotomy, as decades of research have made clear. But the connectome offers a new way to interrogate the relationship between structure and behavior that goes beyond the standard genetic or environmental framings. The fly brain connectome makes it possible for the first time to ask what the relationship is between the physical wiring of a circuit and the behavior that circuit produces, not in one individual but across multiple individuals.
One of the analyses in the FlyWire paper package, led by Gregory Jefferis at the MRC Laboratory of Molecular Biology and the University of Cambridge, compared connectomes from multiple flies to quantify how consistent, how stereotyped, the wiring is from individual to individual. The answer was remarkably consistent. The major circuits, the cell types, the large-scale organization of connections—these are highly reproducible across individual animals. The fly brain is not wired randomly. It is built according to a plan, and that plan is encoded in the genome, expressed during development, and reproduced with striking fidelity. This has a profound implication. If behavior is the output of neural circuits, and those circuits are essentially the same from fly to fly, then behavior in a fly, at least the core species-typical behaviors like courtship, song, feeding, flight, escape from predators, is not a matter of individual variation. It is a property of the circuit architecture itself. The fly does not choose to court; the wiring makes it so.
But the connectome also reveals where variation enters. The fine-scale structure of connections, the exact number of synapses between specific neurons, the precise strength of individual connections—these show more variability between individuals. And this is where experience can intervene. Learning changes synapse strength. Memory is encoded in the modification of existing connections, not typically in the creation of new ones. The connectome, which was made from a single female fly at one moment in time, is a snapshot of a particular brain in a particular state. It is not a universal; it is a reference.
The mushroom body of the fly brain is the region most associated with learning and memory. It is a structure where odors processed by olfactory neurons are associated with positive or negative outcomes and remembered for future behavior. The FlyWire connectome reveals something unexpected about the mushroom body. It also receives visual inputs through pathways that had not been previously characterized. What those visual inputs are doing in a brain region known primarily for olfactory memory is now an active question. The map revealed the connection; the function remains to be established. This is one of the ways a complete connectome differs from a targeted experiment. An experiment asks a question and gets an answer. A connectome asks no particular question. It simply reveals what is there—all of it—and leaves the interpretation to the scientists who come after. It is a resource that keeps generating questions.
There is a harder question underneath all of this, and it would be intellectually dishonest not to address it directly. The connectome tells us how neurons are wired. It tells us which cells talk to which and with what types of signals. Combined with activity data, it tells us how information flows through a circuit and what behavioral outputs that flow produces. It is, in every measurable sense, the most complete structural description of a brain ever achieved. But it does not tell us what it is like to be a fly. This is not a trivial distinction. The so-called "hard problem of consciousness," a phrase introduced by the philosopher David Chalmers in 1995, refers to the question of why any physical process gives rise to subjective experience at all. Why is there something it is like to see red, to feel pain, to anticipate a meal? We can, in principle, describe every neuron in every brain, map every synapse, model every signal. And yet, the question of whether that physical description captures the experience, the felt quality of being the organism in question, remains genuinely open. Whether a fruit fly has any subjective experience at all is a question that the FlyWire connectome cannot answer.
There are researchers who argue that something like consciousness, or at least some primitive form of sentience, some minimal experience of the world, may be present in insects. The evidence is indirect. Flies show behavioral states that resemble mood, exhibit preferences, show signs of something resembling pain avoidance that goes beyond simple reflexes. But behavioral similarity to experience is not experience. And the scientific tools we have for measuring consciousness, which are largely based on behavioral reports and neuroimaging in humans, do not transfer straightforwardly to a creature with a poppy seed-sized brain.
What the connectome does offer is a new kind of foothold on this problem. If consciousness, or any form of experience, has a structural correlate in the brain, then a complete wiring diagram is the place where we might eventually see it. The rich club organization of the fly brain, those densely interconnected hub neurons, resembles, in its organizational logic, the neural structures associated in humans with high-level cognitive functions and, in some theories, with conscious experience. In the human brain, research by van den Hout and Sporns has linked the rich club region specifically to areas engaged in complex behavioral and cognitive tasks rather than specialized peripheral processing. If the same organizational principle appears in the fly brain, does something analogous happen there? We genuinely do not know.
What the connectome does change is the context in which we ask these questions. For most of neuroscience history, the brain has been studied in fragments. This region for vision, that pathway for memory, these neurons for fear. The implicit assumption was that understanding the parts would eventually yield understanding of the whole. The connectome inverts this. It starts with the whole. It gives you every part simultaneously and asks you to understand how they relate. It is a fundamentally different way of approaching a brain. And whether it ultimately illuminates the nature of experience or not, it makes the problem more concrete. The brain is no longer entirely a black box. The structure is visible. The question of what that structure does, what it feels like from the inside to be the creature whose brain it is, remains.
The FlyWire connectome is already reshaping how neuroscience is practiced. More than 50 published studies had used the FlyWire data before the formal Nature publication appeared in October 2024. Within months of publication, that number had grown substantially. Researchers around the world were using the data set through CodeX, the connectome data explorer that the FlyWire team built as an open-access web platform, to identify cell types, trace circuits, plan experiments, and test hypotheses they had not previously been able to address. Anita Devine's description of her lab using the connectome for everything they do captures something real about how a resource of this kind changes a field. The platform is also being used in education. The FlyWire team developed a version of the tool suitable for high school and college students, FlyWire Academy, allowing the next generation of neuroscientists to work directly with real connectome data as part of their training. This matters because the future of connectomics is not one project. It is a field that will produce connectomes for many organisms at multiple developmental stages, in multiple conditions of health and disease. Training scientists who are fluent in this kind of data, who know how to navigate and interpret a connectome the way an earlier generation learned to navigate a genome, is a prerequisite for that future.
The technology pipeline is also moving forward. The same combination of electron microscopy, AI-driven reconstruction, and community proofreading that produced the FlyWire connectome is now being applied to the zebrafish brain. A zebrafish larva has a brain of roughly 100,000 neurons, comparable in scale to the fly, and has the significant advantage that its brain can be imaged while the animal is alive and behaving using light microscopy at near-cellular resolution. A complete connectome of a zebrafish would allow comparisons that are impossible from a fly alone. How does vertebrate brain organization differ from invertebrate? What is preserved across the evolutionary divide between an insect and a fish?
The Microns Project, published in April 2025 after 9 years of work, represents the other frontier: the mammalian brain. The cubic millimeter of mouse visual cortex that the consortium mapped with more than half a billion synaptic connections came with an additional layer of data that no previous connectome had included. The researchers had recorded the electrical activity of many of the same neurons whose structure they then mapped. This meant they could correlate wiring with function. Not just which neuron connects to which, but which ones were active together. Which ones responded to the same visual stimuli. Which ones encoded similar information. A connectome plus activity data is a qualitatively different thing from a connectome alone. It is the beginning of understanding not just the road map, but the traffic.
For the human brain, the challenge remains formidable. At 86 billion neurons and more than 100 trillion synapses, a complete synapse-resolution human connectome would require petabytes, perhaps exabytes, of imaging data, reconstruction algorithms of far greater sophistication than anything currently available, and a proofreading effort that would dwarf every connectome project so far combined. The timeline is not 5 years. It is not 10. But John Ngai, director of the NIH BRAIN Initiative, framed the goal plainly when the FlyWire connectome was published. "Without a detailed understanding of how neurons connect with one another, there is no basic understanding of what goes right in a healthy brain or what goes wrong in disease. The human connectome is not a curiosity. It is eventually a medical necessity."
In the meantime, what the fly brain offers is a model system of extraordinary richness. The fact that 3/4 of the genes implicated in human neurological disease have counterparts in *Drosophila* means that circuits discovered in the fly brain can guide the search for analogous circuits in humans. A circuit for appetite in a fly is not the same as the human neural systems underlying hunger, but they are homologous. They share evolutionary roots. Understanding one deepens the search for the other.
There is something worth pausing on that tends to get lost in the technical achievement. We built a map of a mind. Not a human mind. Not yet. Not even a mammalian mind. A mind the size of a grain of sand, housed in a creature that lives 60 days and eats rotting fruit. But a mind nonetheless. A system that takes in the world, processes it, and produces behavior with a specificity and sophistication that centuries of observation had only partially illuminated. And we built, for the first time, a complete account of how that system is physically assembled.
The philosopher Thomas Nagel asked in 1974, "What is it like to be a bat?" Whether any amount of objective knowledge about a bat's echolocation, its neural processing, its behavioral responses could ever tell us what the bat experiences from the inside. The question remains unresolved. But the connectome is a new kind of answer to a related question: What is the structure that gives rise to experience at all? What is the physical substrate? We cannot yet read consciousness from a wiring diagram. We may not be able to for a very long time, but we have the diagram. We can see how the circuits loop back on themselves, how information from the senses converges on integrating hubs, how those hubs distribute signals that ultimately move muscles and generate behavior. The machinery is visible, and the machinery is, in its way, beautiful. The fact that a brain organized around 78 anatomically distinct regions with 8,453 cell types and 54 million connections generates a creature that navigates, learns, sings, remembers, and sleeps—that all of that emerges from a network small enough to fit on a pinhead—is not a fact that demands less wonder once you understand it. It demands more. Because if a poppy seed brain can do all of this, the question of what 86 billion neurons are doing, what they are generating in you right now as these words form meaning, becomes considerably more interesting. We are each of us a connectome we have not yet read. A wiring diagram of extraordinary complexity, shaped by genetics and experience, running processes we do not understand, producing a continuous experience of being alive that remains, for all the neuroscience of the last century, largely mysterious. The fly brain was not a destination. It was the first legible sentence in a language we have been trying to learn for as long as we have been able to ask the question.
We began with a creature on a piece of fruit. 60 days to live. A brain that fits in the white space between two printed letters. And for most of human history, that brain was as opaque to us as our own. In 2024, a global consortium of scientists, hundreds of researchers across 50 labs, plus volunteers, gamers, paid annotators, and algorithms trained on years of accumulated data, produced the first complete wiring diagram of that brain. 139,255 neurons, 50 million connections, 78 regions, 8,000 cell types. A map you can explore from anywhere in the world through a web browser for free. What followed was immediate: circuits for taste, for learning, for flight, for courtship, for stopping. All rendered suddenly visible, suddenly interpretable. A computational model built from the structure alone correctly predicted how the brain processes different flavors before any experiment was run. The architecture of the whole brain revealed the same deep organizational logic: the "rich club" structure that researchers had found in human brains mapped at far coarser resolution.
The question the map cannot answer remains the deepest one. Whether any of those circuits generates experience. Whether there is something it is like to be a fly navigating a kitchen, singing to a mate, fleeing a shadow. We don't know. We may not know for a long time, but we now have the structure. We have the diagram. And the diagram is the foundation on which every future answer will be built. The brain, this brain, your brain, the brain of every animal that has ever navigated a world it did not make, is not magic. It is machinery. Extraordinarily complex, shaped by hundreds of millions of years of evolution, capable of generating experience that seems to transcend the physical matter that produces it. But machinery nonetheless. Machinery that can be mapped. Machinery that we are beginning, slowly and with great difficulty, to understand. That is where we are. A grain of sand became a map. And the map is already changing what we know.