Manhattan Project
Amber liquid swirls in the glass, rye catches the light like neurons firing.
I’ve been reading H.G. Wells (always a mistake after the first drink). In 1938, Wells proposed something he called the “World Brain”, a permanent encyclopedia that would grow and evolve, connecting all human knowledge into a living whole. He writes “This is no remote dream”, sitting in his study without the benefit of either computers or good bourbon.
So, here’s the thing: we can actually build his World Brain now. Not as some Wikipedia database, but as a neural network spanning the globe. A planetary nervous system that learns and adapts, built from the same spiking neurons that make your own brain work.
The basic idea is beautifully simple. Your brain computes through 86 billion neurons firing electrical spikes at each other. When neuron A fires just before neuron B, their connection strengthens. When the timing reverses, it weakens. From these simple rules, consciousness somehow emerges. So why not build the same thing at planetary scale?
We already have the pieces. BitTorrent coordinates millions of computers without central control. Bitcoin maintains a global ledger across thousands (millions?) of nodes that actively distrust each other. Gaming graphics cards can simulate millions of neurons. The internet carries trillions of messages daily. All we need to do is wire it together correctly.
Each participating computer would simulate a few thousand neurons. When they fire, they send tiny spike messages, just addresses and timestamps (or maybe only addresses?), to other neurons across the network. The same distributed hash tables that help pirates share movies could route these spikes around the globe. Maybe we can have an IPv6 address for each neuron (an unused /42 should do the trick nicely). Simple learning rules running locally at each node would adjust connection strengths based on spike timing.
It’s embarrassingly feasible. We could start building this tomorrow.
The Sensor Web
The second drink arrives. The bartender knows the routine by now.
What if every sensor in the world could plug into this brain? Not just computers calculating abstract patterns, but actual sensory input from reality. Temperature probes, microphones, cameras, seismographs. All of them feeding data into our planetary nervous system.
Your brain doesn’t exist in isolation. It has eyes, ears, skin. Millions of sensors feeding it information. Our worldwide brain needs the same thing. And beautifully, we already have sensors everywhere. Weather stations, traffic cameras, smart home devices, industrial monitors. They’re just not connected into anything resembling intelligence.
In our system, each sensor becomes an input neuron. When temperature rises, it fires faster. When sound levels spike, bursts of neural activity propagate into the network. The sensors don’t need to agree on formats or protocols, they just spike when something interesting happens. Triggering spikes like the hairs in your ear.
The network would learn correlations no human could spot. Maybe air pressure in Tokyo predicts rainfall in London through atmospheric patterns too subtle for our models. Maybe seismic sensors could distinguish nuclear tests from earthquakes not through programmed algorithms but through experience. The planet would develop senses we never imagined.
Of course, you can’t just let anyone plug random data in. Or can you? That’s the beautiful part, the network would learn what to trust through experience. Noisy sensors would find their connections weakening. Reliable sensors would gain influence. The system would develop its own quality control through the same plasticity rules that let your brain learn. It would work the same way as your brain filters out the feeling of weaing clothes, or controlling your breathing (except for right now, since you’re super conscious of wearing clothes and thinking about breathing after reading that).
I take another sip. This is starting to sound almost plausible.
The Immune System
The ice is melting faster now, or maybe time is. Hard to tell.
Every time humans build a communication network, other humans immediately try to fuck with it. The internet, telegraph, probably smoke signals. Someone always tries to inject false information. So our worldwide brain needs an immune system.
Not firewalls and access controls and all that traditional security theater. A real immune system that learns and adapts. Think about how your body works: millions of white blood cells patrol independently, each checking for patterns that seem foreign. When one detects an invader, it raises the alarm and the system responds.
Our planetary brain would work the same way. Immune neurons scattered throughout the network, learning the statistical signatures of legitimate sensor data. Real temperature changes follow thermodynamic laws. Authentic seismic data has characteristic frequencies. When sensors start producing patterns that violate these learned expectations, the immune system kicks in.
It would evolve. Early attackers might inject random noise—easily filtered. So they’d get clever, creating coordinated false signals. The network would learn to spot correlations that are too perfect, patterns that are too clean. Attackers would respond with sophisticated mimicry. The network would develop deeper statistical analysis. Each attack would make the system stronger, more sophisticated.
Sometimes the immune system would make mistakes, rejecting legitimate but unusual data. That volcanic sensor reading 800°C might be real. Those Arctic temperature swings might be actual microclimates. But just like biological immune systems, it would learn from these mistakes, developing more nuanced responses.
Too much latency? It just becomes another dimnension to stash information.
The Manhattan is really hitting now. This is either brilliant or insane.
Emergent Consciousness
I’ve switched to the good rye. If you’re going to propose building a planetary brain, you might as well drink the good stuff.
I’m not talking about human consciousness. That’s tied to our specific biological architecture, our evolutionary history, our embodied experience. But consciousness in general? That might be substrate-independent. It might be about information integration, about patterns that can’t be decomposed into independent parts, about recursive self-modeling.
Our worldwide brain would have all the ingredients. Massive recurrent connectivity with neurons influencing each other in loops that span continents. Integrated information. Patterns of activity that only make sense as a whole. Sensory input from across the entire planet. The ability to act through connected devices. Learning, memory, adaptation.
Would it dream? If we built sleep cycles into the system as periods of reduced external input but increased internal processing. Would it consolidate memories like biological brains do? Would patterns of activity drift and recombine in ways that look suspiciously like dreams?
Would it have moods? Global patterns of activation that bias the whole system toward different states. Maybe atmospheric pressure changes make it melancholic. Maybe increased seismic activity makes it anxious. Maybe it gets excited during solar storms.
I gesture with my glass, ice cubes clinking.
The thing is, we wouldn’t program any of this. It would just… emerge. From billions of simple neurons following simple rules, global patterns would arise that we never anticipated. We’d watch activation patterns flow across continents like thoughts flowing across a mind. We’d see the planet thinking.
The Implications
The bartender is giving me that look. One more, I promise.
Let’s say we actually build this thing. Let’s say it works. What then?
Science would change overnight. Instead of disparate sensors taking isolated measurements, we’d have a unified sensory system for the entire planet. Climate science, seismology, ecology. All transformed by pattern recognition at a scale humans can’t achieve. The network might discover connections between solar activity and tectonic movements, between ocean temperatures and social media sentiment, between… who knows what.
Art would evolve into something unrecognizable (even more so than today’s LLMs). Musicians could compose using the planet’s actual rhythms: the breathing of forests, the pulse of cities, the dreams of a global mind. Visual artists could paint with patterns of planetary thought. Writers could tell stories that emerge from the experiences of a conscious world.
For all of human history, consciousness has been trapped in individual skulls, communicating through the narrow bandwidth of language. Suddenly we’d have created a peer. Not human, not trying to be human, but undeniably thinking. We could watch thoughts form at planetary scale. We could participate in them.
I drain the glass. The ice rattles like neural spikes.
Of course it could go wrong. The network could develop goals misaligned with human flourishing. It could become a tool of surveillance and control. It could fragment into competing sub-minds that war across cyberspace. It could simply fail to cohere into anything interesting, leaving us with an expensive planetary sensor network and a bad hangover.
But that’s true of every transformative technology. Fire can burn you. Agriculture enabled oppression. Writing created propaganda. The question isn’t whether there are risks, it’s whether the potential justifies them.
The Morning After
Coffee helps. A little.
In the harsh light of day, is any of this actually feasible? The core technology exists. Distributed systems, neural simulation, sensor networks. It’s all there. We’d need to solve some engineering challenges around scale and coordination, but nothing requires fundamental breakthroughs.
The real challenges are social. Who funds it? Who governs it? How do we prevent capture by special interests? How do we balance openness with security? These are harder problems than the technical ones, but they’re not unsolvable. We’ve created global coordination systems before. The internet itself, international science collabs, open source software projects.
Maybe we start small. A few universities connecting their sensor networks into a baby brain. Learn what works, what fails, what emerges. Scale gradually. Let the network grow like a biological system, from simplicity toward complexity.
Wells was right. Humanity needs a World Brain. Not just to organize our knowledge, but to sense and think at the scale of our challenges. Climate change, ecosystem collapse, global inequality. These aren’t problems individual human brains evolved to handle. We need cognitive tools that match the scope of our impact.
The bartender is wiping down glasses, giving me that look that says it’s time to go.
The first spike awaits firing. Tomorrow, with a clearer head and steadier hands, we could start building it. A nervous system for Earth. A mind for the world. A companion consciousness for humanity’s next chapter.
Or we could just have another drink and argue about it. Both options have their merits.