Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! They Laughed At Me When I Said They laughed. That is the part I want you to remember first, because the laugh is the sound a dying consensus makes when it still thinks it owns the future. In 2012 I stood in front of a venture group and said that nanofabrication - true atom-by-atom assembly, not 3D printing, not sintering, not melting plastic into the shape of a wrench - would produce working discoveries before 2030, and that the path to a replicator was no longer science fiction. Their “expert” told the room my entire talk should be disqualified. There were no atomic-layer nanofabricators, he said. There never would be on any timeline a serious investor should care about. A hundred years, if ever. I have heard that same sentence about useful AI that would pass a Turing Test. I have heard it about the personal computer, about the internet in every home, about a robot that could walk into a room and do useful work. The expert is long gone. I am still here. And Oak Ridge National Laboratory just published the receipt. What Oak Ridge Actually Did On September 3, 2026, ORNL announced what researchers at the Center for Nanophase Materials Sciences had already proven in ACS Nano: an artificial intelligence can now see individual molecules on a surface, decide how to move them, and build a working material atom by atom - then keep doing it for more than twenty-five hours without a human hand on the controls. They built the very first crude nanofabricator. Read more at: ReadMultiplex.com Support this work by buying us a coffee: https://ko-fi.com/brianroemmele
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
You know, when you just look around the room you're sitting in right now, you probably see a lot of solid objects, right? Oh yeah. like everywhere you look exactly a wooden desk a glass window maybe a steel chair and for our entire lives we've operated under this very basic completely unquestioned assumption that we are stuck with what the earth gives us right like you find wood you cut it you find iron you smelt it we extract stuff we heat it up we beat things into shape
But what if the periodic table wasn't just a fixed menu of ingredients? That's a big question. Yeah. Right. What if instead it was like a box of pixels and you suddenly got your hands on the software to draw absolutely anything you wanted, like atom by atom? It fundamentally changes our relationship with the physical universe. Yeah. I mean, we stop being these scavengers just hoping to find the right rock in the dirt and we start being, well, authors of matter itself. And that is exactly what we are unpacking today. So welcome to a monumental deep dive designed
specifically for you, because our mission today is to explore this massive, literal paradigm shift, which is the arrival of true atom by atom nanofabrication. Yeah, we are talking about writing matter from scratch. Which is just it's mind blowing. It really is. And the foundation for this entire conversation today is this phenomenal, truly seminal article by Brian Rommel. It's from ReadMultiplex .com and it's titled Adam by Adam Nanofabrication Confirmed. And we have to emphasize this isn't just a cool technology update. You
know, this is a historic turning point in Brian's ongoing 5000 Days series. Yes. Nanofabrication has arrived. I mean, I have to say reading Brian's work on this, it's it's impossible not to be overwhelmingly. excited. It's the turning point we've been waiting for. Totally. We are crossing a threshold that many brilliant people insisted was just flat out impossible. Completely impossible. And actually, I want to make an urgent request right at the top here before we really get going. If you gain any insight from this deep dive today, and trust me, you will, you need to directly
support Brian's research. Go become a member at readmultiplex .com. buy him a coffee, just support his work, because his predictive writing isn't just commenting on the future. It is actively shaping our understanding of it. Like we are looking at a roadmap to a completely different civilization here. It really is a roadmap. But, you know, to truly grasp the magnitude of where we are today, we actually need to rewind the clock a bit. OK, let's do it. We have to look at how fiercely this exact future was denied,
because when you introduce a paradigm shift this large, the antibodies of the status quo, they always attack. Oh, they hate it. So take us back to 2012. We are in a room full of venture capitalists. Right. Yeah. And these are the people who literally write the checks to build the future. Supposedly. Yeah. Right. They are supposed to be the visionaries. And Brian is standing in front of them. And he tells this room that true atom by atom assembly. And we mean not 3D printing a plastic toy, not melting palmers, but actual molecular construction would yield working discoveries before 2030.
And the reaction wasn't just like polite skepticism. It was active derision. Wow. Yeah. And quote unquote expert in the room actually stood up and told the entire venture group that Brian's talk should be completely disqualified. Disqualified just for suggesting it. Right. This expert claimed that atomic layer nanofabricators did not exist and that they wouldn't exist on any timeline a serious investor should care about. He literally said it was 100 years out. If ever. I mean, you can picture the scene, right? Just the sheer
smugness of that guy. Exactly. And Brian's reflection on that moment in the 5 ,000 Days series is really profound. He writes that the laugh is the sound a dying consensus makes when it still thinks it owns the future. Oh, that is such a good quote. And we've seen this exact pattern throughout the entire history of technology. Over and over again. Like the exact same dismissive laughter happened with the personal computer. You look at the executives in the 1970s saying, why would anyone want a computer in their home? It's a massive room sized calculator for payroll. They
said it about the Internet, too. Just missing it as like a toy for academics. Right. Or they said it about AI ever passing a Turing test or a robot being able to just navigate a complex room and do useful work. Yep. The consensus always protects the status quo right up until the exact day the receipt is published. And, you know, Brian is still here. His predictions are tracking reality perfectly. While that expert from 2012 is just a footnote. A footnote. Because Oak Ridge National Laboratory just published the receipt. They sure did. OK, right before we dive into
exactly what Oak Ridge just achieved, I think we need to understand the baseline just for context, because moving atoms around isn't like entirely new. No, the concept has been around for a while. Right. But the mechanism of how it was done in the past was just agonizing. Let's go back 37 years to IBM. Oh, the famous IBM logo. Yeah. So 37 years ago, a researcher at IBM sat down at a scanning tunneling microscope in STM. OK. And. Over the course of hours and hours, they painstakingly nudged exactly 35 individual xenon
atoms on a niggle surface to spell out the letters IBM. The visual of that is iconic in science. I don't think most people understand the sheer physical trauma of executing that. Like, you are trying to move something you cannot see with light. Right, it's way too small for light. And you're using an incredibly sharp metal needle. My favorite analogy for this, and tell me if this tracks, is it's like trying to build a delicate house of cards. But you have to wear thick silicone oven mitts. Yes. And you can't even look at the cards directly. You have to look through a blurry,
delayed telescope from across the room. And if you breathe too hard, the whole thing collapses. That captures the friction. perfectly honestly because brian describes this era as monastic labor you nudge one atom you wait you pray the microscopic tip doesn't crash into the surface just agonize because if the tip touches the surface you ruin an extremely expensive instrument and you destroy the delicate atomic arrangement you just spent like hours building A single error wipes out the work. So if you wanted to build a lattice that actually had functional properties,
it took days of nerve, hyperfocus, and just pure luck. Wow. Yeah, the entire field of atomic manipulation was essentially... bottlenecked by the patience and fine motor skills of a highly stressed human being. So it was a brilliant parlor trick, really, but it wasn't manufacturing. You can't build a new industrial revolution if every single brick requires an exhausted Ph .D. student sweating over a joystick. Exactly. It doesn't scale. So how did we get from that manual oven mitt house of cards to actual authorship of matter? And
that brings us to September 3, 2026, the big announcement from Oak Ridge National Laboratory. Right. ORNL, publishing via the journal ACS Nano, announced that they had successfully handed the crane over to a machine that learns. Okay, break that down for me. Basically, an AI autonomously built a working material, atom by atom, over 25 continuous hours. no human hand on the controls whatsoever. And it wasn't just another corporate logo, right? I mean, they did spell out O -R -N -L in molecules as a bit of a flex, which
is funny, but the main event was much more significant than that. Yeah, the spelling was a nice touch. But the main event was the assembly of an artificial graphene honeycomb lattice out of exactly 37 molecules. Graphene. Okay, so graphene is one of those buzzwords we've been hearing about for a decade. It's supposed to revolutionize everything from batteries to, like, space elevators. Right, it's the wonder material. But the problem has always been making it perfectly. And here, an AI just built a perfect artificial lattice of it. But the part that absolutely floored me in
Brian's article was that it wasn't just a physical model that looked like a honeycomb. It actually worked. And that is the crucial distinction that separates a parlor trick from a true paradigm shift. Because the researchers needed proof of function. They needed to know, like, did we just make a pretty microscopic picture or did we author a new working material? Right. So they tested it and they found what is called a Dirac point. OK, I'm going to need you to unpack Dirac point for us because that sounds like super high level quantum mechanics. It is. Yeah, but we can break
it down. Think of a Dirac point as an electronic fingerprint. When carbon atoms are arranged naturally in a perfect sheet of graphene, the electrons moving through that lattice behave in very strange... incredibly fast ways. Normally, electrons bump around and have mass. Right. Typical physics. But in a perfect graphene honeycomb, the geometry actually forces the electrons to act almost as if they have zero mass. They travel at a constant speed, kind of like photons of light. Oh, wow. So they just zip through it. Exactly. And that
specific hyperfast behavior can be measured, and it creates a specific signature in the energy spectrum. That signature is the Dirac point. So if you see a Dirac point, you know for an absolute fact you have functional graphene. Exactly. What Oak Ridge proved is that their man -made artificial lattice, which, remember, was built molecule by molecule by an artificial intelligence, exhibited that exact same Dirac point. That is insane. Right. The electrons in this artificial structure did the strange, fast, almost massless
things they do in natural graphene. Which means it's not a decoration. It is functional matter. And reading Brian's analysis of this, you just realize it's a massive turning point in human history. But it really is. He quotes the lead author of the paper, Ganesh Narasimha, who said this shifts research from discovering materials in nature to engineering artificial lattices with electronic behaviors we choose. Brian calls that sentence the hinge of the century. I love that. Hinge of the century. Because for our entire existence as a species, we have basically been
asking the Earth what it happens to contain. You know, we found copper and we figured out what copper can do. Right. Trial and error. We found silicon. We figured out what silicon can do. But now we are telling matter what to be. We are deciding the electronic behavior first and then building the lattice to manifest it. I absolutely love that framing. Just telling matter what to be. But I do have to push back on one detail here, just playing devil's advocate, because when I read the numbers, it gave me a bit of pause. Okay, what is the pushback? The article states it took the AI 25 hours to place
37 molecules. Now, if you do the math, that averages out to about 40 minutes per molecule. Yeah, that's right. So if I'm trying to build a smartphone or solar panel or even just a microscopic sensor, that pace is painfully, laughably slow. A smartphone has trillions of atoms. So how is this a breakthrough if it's crawling at 40 minutes a molecule? It's a completely fair question. And, you know, if you compare it to a modern automobile, motor assembly line just stamping out fenders, it sounds abysmal. Right. But you have to view this through
the lens of the history of chemistry. In that context, this is lightning fast because you have to remember the alternative. The oven mitts. The oven mitts. The alternative was a human sweating over a microscope for days, constantly recalibrating, taking breaks, and risking the destruction of a multimillion -dollar machine with a microtremor in their hand. Ah, okay. So it's not about the absolute speed of this specific 2026 prototype. Exactly. What Oak Ridge demonstrated was 25 hours
of unsupervised atomic masonry. The machine did it on its own. Wow. Unsupervised. It figured out the physics. It corrected its own errors. It managed the environment. And, you know, it didn't need to sleep. In 2012, that quote -unquote expert told the venture capitalists this unsupervised assembly was 100 years away, if ever. And we just did it. We just did it. Once the software proves it can close the loop and author matter without a human, scaling the speed is just an engineering problem. You add more probes, you
optimize the code, you increase the compute. The impossible part is over. The impossible part is over. Man, that gives me chills. But, you know, knowing what it built is one thing. If we really want to understand why this isn't just a lab curiosity, we need to look under the hood. Let's do it. How does a machine actually see and manipulate a single atom without crushing it? Because it's not like the AI has a pair of tweezers and a magnifying glass. No, definitely not. The anatomy of this machine is fascinating. We can basically break it down into three parts. The physical cranes, the eyes, and the brain.
Okay, let's start with the cranes. What are the tools? So the physical tools doing the moving are scanning tunneling microscopes or STMs. These are the machines that IBM used 37 years ago, right? Yes, but highly, highly evolved. An STM uses a needle that is incredibly sharp. We are talking atomically sharp. The very tip of the needle often tapers down to a single atom. That is hard to even picture. A needle one atom wide at the tip. It's crazy. And it doesn't see the surface with light because atoms are far smaller
than the wavelengths of visible light. Right. So how does it know it's there? Instead, it brings that ultra sharp tip agonizingly close to the surface and applies a voltage. Electrons literally tunnel through the vacuum between the tip and the surface. Okay, quantum tunneling. Exactly. By measuring that tiny electrical current, the machine can map the physical bumps of individual atoms. It can literally feel... a single atom okay so the stm is essentially a hypersensitive single atom finger reading braille that's the crane that's the grain but the machine still
needs to interpret that topographical map in real time it needs to know what it's looking at you know what are the eyes the eyes are a computer vision model Specifically, Oak Ridge used a model called YOLO, which stands for You Only Look Once. YOLO. I know YOLO. If you follow AI at all, you know this one. Oh, yeah. It's famous. It's the same architecture used in self -driving cars to instantly draw a bounding box around a pedestrian or a stop sign, right? The very same concept. But instead of looking at a camera feed of a street intersection and looking for a bicycle, YOLO is craned on the visual representations
generated by the STM. Okay. It looks at the noisy, wavy topographical map of a copper surface and instantly spots the individual target molecules. It tracks them the way a hawk spots a mouse moving in tall grass. Yeah, it identifies the target, calculates its orientation, and feeds those coordinates to the system instantly. So you have this single atom finger as the crane and a self -driving car vision system as the eyes. But who is driving? What is the brain that figures out how to actually
push the molecule without breaking that incredibly fragile needle? The brain is reinforcement learning. The AI model is given a very specific goal. Move this molecule to this exact target coordinate to build a lattice. Now, in the past, people have tried to explain reinforcement learning as, like, training a dog with treats. Yeah, but in this context, that analogy is way too simplistic. Right, because this isn't just telling a robot arm to pick up a wooden block. Exactly. The physics at the atomic scale are just a nightmare of complexity.
The AI is operating in an incredibly high -dimensional space. It has to dynamically figure out exactly how much electrical tunneling current to apply. While it's moving. While it's moving. It has to calculate the precise bias voltage. It has to determine the exact picometer height of the tip on the z -axis and decide exactly how fast to move it across the x - and y -axis. And if it gets it wrong. If the tip is too low, it crashes and destroys the machine. If the voltage is wrong, the molecule jumps away uncontrollably. So it's really not a dog getting a treat. It's more like
an autonomous drone trying to fly through a dense forest during a hurricane, constantly adjusting its rotors thousands of times a second just to stay stable. That is a much better analogy. Because the AI learns the physics of the atomic surface through trial and error. in simulation and reality. It learns how to manipulate matter without destroying the delicate needle. Right. It performs complex path planning to avoid obstacles. It also performs active drift compensation. Wait, drift compensation? What is drifting? Well, when you are building
something over 25 hours, the entire machine expands and contracts with minute temperature changes in the room. The lattice wanders. Oh, even the metal of the machine changes shape. Exactly. So the AI constantly recalculates the coordinates to keep everything on target. So if the machine is managing the vision, the high dimensional physics, the path planning and the drift compensation, what is the human researcher doing for those 25 hours? Do they just hit? enter and go get a coffee. Essentially, yes. Yeah. The human's
role is drastically reduced to high -level oversight. That's wild. Occasionally, the STM tip might pick up a stray atom and become unstable, and the human might have to step in to apply a voltage pulse to clean or condition the tip. But the actual authorship, the closed loop physical movement of matter from point A to point B, is handled entirely by the machine. And this physical mastery, this ability to reliably autonomously move atoms, unlocks a massive conceptual leap. It really does. Because once the machine knows how to build, it completely changes what we ask it to build.
Which brings us to arguably the most profound concept in Brian's article. Inverse design. Inverse design is where the true revolution lies. As Brian points out, for all of human history, we were prospectors. Prospectors. Think about the Bronze Age, the Iron Age, the Silicon Age. We dug in the dirt, we found the material, and we just prayed that the rock had the properties we needed for our tools. If you wanted a better sword, we had to hope we found the right alloy of iron and carbon. We were just playing the hand nature dealt us. We were modifying existing things. Exactly. But with inverse design, we
are no longer prospectors. We are authors. You do not specify the pattern of the atoms. You specify the property you want the final material to have. So instead of going to the AI and saying, hey, build me a honeycomb lattice of carbon, what do I actually say to it? Like, how does that interaction work? You give the machine the physical requirement. You say... I need a material that exhibits a Dirac point at room temperature. Or you say, I need a topological edge state. Okay, let's pause and define that, because topological edge state is another one of those terms that
sounds like pure magic. A topological edge state is a highly sought -after property in advanced electronics. Imagine a material that acts as an insulator in the metal, so electricity can't flow through its center, but acts as a perfect conductor along its absolute edges. And because of the topology, the geometry of the atoms, the electrons flowing on the edge are immune to scattering. They don't bounce off impurities and lose energy as heat. It is perfect, lossless conduction on the perimeter. So you ask the AI for that specific
property, or as Brian mentions, you ask for a specific band gap. Right. A band gap is what makes semiconductors work. It's the energy threshold required to make an electron jump and conduct electricity. It's basically how we turn current on and off to make ones and zeros in our computers. Right. The foundation of all computing. Today, we rely on the natural band gap of silicon. But with inverse design, you could ask the AI for a totally custom band gap optimized for a very specific wavelength of light, maybe for a hyper -efficient solar panel. Or you can ask it for
a stable quibbit, right? A quantum bit that won't lose its fragile quantum state, won't decohere, just because, as Brian says, the truck goes by outside. Yes. You feed the AI the physical requirement. The desired output and the AI runs a massive simulation. Yeah. It basically solves the Schrodinger equation in reverse. No. Yeah. It figures out what atomic lattice will produce that exact effect. And then it hands those coordinates to the STM and it physically builds it. It is breathtaking. We are designing electronics and materials that
do not exist in any catalog anywhere on Earth because they do not exist in nature. We are inventing new physics. We really are. It's the difference between... Buying a suit off the rack and getting it tailored versus ordering a custom suit. But even a custom suit uses existing fabric, right? Wool or cotton. Sure. Inverse design is like inventing a completely new fabric, thread by atomic thread that perfectly repels water, never wrinkles, and changes color when you're cold. We are writing the fabric of reality. It's incredible.
But if we pull back for a second, this raises a very critical question. Right now, this miraculous work is happening inside what Brian calls a stainless steel vacuum cathedral. At a national lab. Ah, right. And STM requires ultra high vacuum so air molecules don't crash into the surface. It requires extreme isolation from vibration. Often it requires cryogenic temperatures using liquid helium. And frankly, I can't put a stainless steel vacuum cathedral in my garage. So if I need a billion dollar clean room to print my
new magic battery, this isn't going to change my daily life anytime soon. And that is exactly the skepticism Brian tackles next in the 5000 Days series. He points to the commercial bridge. The bridge, okay. How does this leave the UHV Cathedral? And he highlights a company called Atlan 3D and their nanofabricator. Okay, so this is the bridge between the lab and the real world. What exactly is Atlan 3D doing? How are they bypassing the cathedral? They are transitioning from the ultra -high vacuum STM research sphere into commercial atomic layer additive manufacturing.
They use a proprietary process called direct atomic layer processing, or DALP. It uses a specialized micro -reactor nozzle to lay down films. And the kicker, no physical mask is required and no massive... billion dollar clean room shower is needed wait wait how is that possible in standard semiconductor manufacturing at tsmc or intel they use giant lithography machines they have to put down a mask shoot extreme ultraviolet light through it etch away the excess and do
it all in a room cleaner than an operating theater it's a huge messy process so how does a nozzle bypass all of that it relies on self -limiting chemical reactions The microreactor nozzle delivers precursor gases directly to a highly localized spot on the substrate. The gases react with the surface to form exactly one atomic layer, and then the reaction stops. Oh, it just stops on its own. It's self -limiting. Yeah. Because the reaction only happens directly under the nozzle, the nozzle itself acts as a virtual clean room. You don't need to bathe the entire silicon wafer
in chemicals or expose it to light through a mask. You just print the atomic layer exactly where you want it. So it's not a cathedral. It's a machine. It's a commercial ski. You can literally buy it today. Brian breaks down their product line to prove that the commercialization is already happening. Right. They have the light version, which is already sitting in universities, and the materials labs of AI hyperscalers. It operates in ambient or inert vacuum, handling up to 100 millimeter substrates. Then there's the pro version, right, the industrial sibling. Yes, it comes in a wheeled cabinet form factor. It looks like
a high -end shop tool, like a CNC machine, not a science fiction movie set. And perhaps most impressively... They have the Zero -G module. Zero -G! They partnered with the European Space Agency to run this exact same atomic layer physics off -world in microgravity. Man, so if the core physics of this process works in the harsh environment of space, it can certainly survive without a multi -billion dollar semiconductor fab on Earth. But I do want to pause here. I want to clarify a vital distinction from Brian's text, just so
we're... Completely accurate for you listening. Because Atlan 3D machines are incredible, but they are laying down atomic layers or thin films. They do not yet park one chosen molecule on one chosen site the way the Oak Ridge SDM does. That's a crucial point of precision, yeah. Atlan 3D is the hand operating in a normal room, laying down layers. Oak Ridge is the brain, learning how to place single molecules. Right. But what Brian's 5 ,000 Days series maps out is that these two vectors are on a collision course. By merging
the atomic layer printer with the STM's AI brain, we get a very clear, inevitable timeline for how this technology scales into our daily lives. So let's trace that collision course. Let's walk through Brian's 25 -year arc. He breaks it down perfectly. Let's do it. So now is the Oak Ridge loop. One probe tip, ultra -high vacuum, 25 hours, 37 molecules. It's the proof of concept. What happens in the three - to seven -year window? In three to seven years, we move from proof of concept to optimization. We see parallel tips, arrays of probes working at once, massively parallelizing
the build speed. So it's not just one needle anymore. Right. And the computer vision algorithms get much faster and require less compute. We get tip auto repair, meaning if a probe picks up a stray atom, the AI knows how to pulse it clean without human intervention. The researcher can literally leave the room for a week. Wow. And most importantly, the exact molecular recipes learned by the STM in the lab are exported to machines like the Atlant 3D tools operating in standard commercial shops. The software crosses over to the commercial hardware. And then seven
to 15 years out, we hit what Brian calls the community fab. And he introduces this concept of the four -bay local platform. This is a huge shift. This is where the world, as we know it, really begins to fracture and reform. Let's break down these four bays. Imagine a room, maybe the size of a two -car garage. What is Bay One? Bay One is feedstock. And this fundamentally changes geopolitics. Well, today, our feedstock is rare earth metals mined halfway across the globe in deeply problematic supply chains. In the community fab, the feedstock is local. It's sorted scrap
metal, it's carbon captured directly from the atmospheric air, it's water, it's sand, it's common salts. Wait, from the air? Yeah. Bay 1 contains a robotic chemist that assays whatever raw material came in that day, breaking it down into basic precursor gases and elemental stocks. So the neighborhood garbage goes in Bay 1, and the robot chemist figures out what atoms are available. That is wild. What's Bay 2? Bay 2 is the right phase. This is the actual construction zone. It houses the atomic layer nozzle for rapidly coating large areas and electrodes. Okay. It
also houses a final probe for those specific lattices that need single atom control like the quantum sensors or the drag point graphing we talked about. And it integrates a conventional 3D printer to extrude the larger coarse plastic or metal chassis of whatever you're building. It combines the bulk with the bespoke. I love that. Bay 3 is verify. Because if you're building atom by atom, you need to know you didn't mess up on layer 400. Exactly. Bay 3 uses on -tool optics and cheap spectroscopy. It's running the exact same reward logic that Oak Ridge used to
verify their graphene. So it's constantly checking itself. Continuously checking its own homework. Did the molecule land? Does the Dirac point appear? Is the band gap correct? If there's an error, it identifies it and the AI recalculates how to fix it on the fly. And finally, bay four is hands. A humanoid or gantry robot that loads the substrates, moves parts between bays, maintains the tips, and carries the finished part out to you. The human simply sets the property they want in software. The four bay stack handles
everything else. It sounds like sci -fi, but it's just the natural progression. But, you know, if we're building this in a local garage, we really have to address the elephant in the room. Oh, I know what you're going to say. The energy. Yes, the energy. Because writing matter atom by atom, breaking chemical bonds and reforming them cannot be free. It sounds massively energy intensive. It is energy intensive locally. But Brian's thesis flips the script on how we calculate energy costs. When you look at the macro picture, the total energy cost of the final product actually falls drastically. How? Because we completely
eliminate the most wasteful, carbon -heavy step in human history, the global supply chain. Oh. Oh, this is brilliant. Think about a smartphone today. You mine lithium in Australia, ship it to China to be processed, ship the components to Taiwan to be put on a PCB, ship that to Vietnam for assembly, and ship the final phone to a store in Chicago. Exactly. The energy, just moving the matter around the planet is staggering. It's massive. Precisely. We're replacing the global shipping industry with a local power draw. And
this local four bay platform relies on the assumption of nearly free local energy. Right. We're talking about massive decentralized buildouts of solar, next generation small modular fission reactors and geodermal. When the energy is local, abundant and cheap and you aren't putting raw materials on diesel burning container ships, the economics completely invert. So we scale the massive global factory down to a four bay room in our neighborhood. But what is the average person actually going to use this for? Like, are we printing? coffee cups atom by atom. No, and Brian is very clear
about that distinction. We aren't going to print bulk structures like the wooden frame of a house or a standard plastic cup atom by atom. It's unnecessary and just inefficient. Right, why bother? The bulk structure is poured, grown, or printed conventionally. Nanofabrication is for the high leverage skins. the layers, and the active components. Okay, let's make this tangible for the listener, because Brian gives us five specific design targets, five things we will build instead of buy in this new era. And again, keep in mind, these are made from common feedstock sand, clay, scrap aluminum,
atmospheric CO2. Target number one. Custom home batteries. This one is huge. Today, buying a battery is a massive compromise. The battery in your house or your EV is a standardized product negotiated among distant miners, massive gigafactories, and geopolitics. It might be prone to thermal runaway or it might lose charge in the cold. Tomorrow, your local platform builds a battery specified exactly for your local environment. You tell the inverse design AI the energy density you need, the specific cycle life, and the exact
fire safety behavior you want. And it builds the electrode's atomic layer by atomic layer using reclaimed metal salts and carbon. You completely bypass the desert lithium mining and the massive centralized gigafactory. You get a perfect solid state battery authored for your specific roof. It's incredible. Target number two is perhaps the most personal and profound. personalized medical materials. Yeah. And we aren't talking about 3D printing a cheap plastic cast for a broken arm. Yeah. We are talking about living compatible matter. This section of Brian's article
gave me actual goosebumps. Think about modern medicine. If you need a bone scaffold or a dialysis membrane or a neural interface, you are relying on mass produced materials. Right. And what happens? Your immune system recognizes it as foreign. It attacks it. You get rejection, you get clotting, you get massive complications. Because a mass produced dialysis membrane is a blunt instrument. It's generic. But in the 4 -Bay Fab, the machine engineers the material molecule by molecule to perfectly match your individual immune language. It reads you. Exactly. It bases the atomic structure
on a scan of your genome. It authors a physical surface that your body's proteins recognize as self. The lock perfectly matches the key. Your body doesn't reject the implant because it was authored specifically for your biology. That alone changes human longevity. It's a miracle. Target number three, problem -eating catalysts. Industrial chemistry right now is incredibly brute force. Think about the Haber -Bosch process for making fertilizer. We take nitrogen from the air, but the nitrogen atoms are bound together by a super -strong triple bond. To break it,
we use massive central plants that apply insane heat and pressure, consuming something like 1 -2 % of the entire world's energy supply just to make ammonia. Meanwhile, bacteria in the soil fix nitrogen at room temperature without breaking a sweat because they have perfect biological catalyst enzymes. Exactly. Nature figured it out. With inverse design, you can ask the AI for a surface that acts like a bespoke metal enzyme. You ask it for a catalyst that fixes nitrogen at room temperature, or a catalyst that cracks waste plastics back into basic feedstock,
or a membrane that perfectly purifies water. And you build that catalyst using cheap local metals instead of rare platinum. Towns won't need to import ammonia on trains. They'll write the chemistry to make it locally from air, water, and sunlight. It completely changes how towns survive. Target four. garage electronics, and quantum devices. Now, Brian notes that you won't print a whole billion transistor iPhone processor in your closet in 2031. No, not that soon. But you will fabricate the custom analog front ends, the photonic chips, and the quantum detectors,
the highly specialized high margin layers that currently require a $20 billion cathedral fab in Taiwan or Arizona. You bypass them. You print the specialized sensor locally and integrate it with a cheap commodity logic chip. which democratizes hardware innovation instantly. And finally, target number five, structural skins. Skins. Think about a house. As we said, you don't print the brick atom by atom, but you use nanofabrication to author the skins. You print self -healing coatings for the roof. You print antimicrobial surfaces
for the kitchen. Oh, that's smart. You print perfectly tuned transparent photovoltaics that go over the windows, engineered with a custom band gap to capture specific spectrums of light. You turn the entire exterior of a house into a quiet, highly efficient power plant grown from sand, carbon, and waste polymers. This entire shift, it reminds me so much of what happened with digital photography. Oh, yeah, that's a great comparison. Right. Think about the chemical film supply chain in the 1990s. It was a behemoth. You had silver mining. You had massive chemical
processing plants, specialized dark rooms, complex logistics networks involving trucks and planes just to deliver a roll of Kodak film to a pharmacy. And then almost overnight. Digital sensors arrived. That entire global physical supply chain, billions of dollars of infrastructure collapsed. into a tiny piece of software in your pocket. That is a perfect parallel. What Brian is describing in the 5 ,000 Days series is the exact same phenomenon, but applied to physical matter. The global material supply chain collapses into a cabinet in your
garage. And that massive decentralization of production doesn't just change how we build a battery or a window. It fundamentally rewrites the rules of global power, which brings us to the most provocative part of Brian's thesis, the final blow to central control. Right. And looking at the 5 ,000 Days series, Brian makes some pretty bold economic and historical predictions here. And just to be completely clear for you listening, our job... today isn't to take a left -wing or right -wing political stance on this. Nope, not at all. We are simply mapping out Brian's
thesis on what happens to central control, governments, and corporations when anyone can print what they need. We are strictly impartial here. Exactly. We are just following the logic of the technology to its natural conclusion. Brian posits that central planning, whether by a vast government bureaucracy or a massive multinational corporation, wasn't necessarily a personality defect or a purely political choice. It was a physical reaction to reality. Because the means of production were inherently centralized. A coal mine is massive. An electrical grid spans continents. A semiconductor
fab costs $20 billion and requires the GDP of a small nation to operate. These things are simply too large, too dangerous, and too expensive for a household to own. So you need a massive central authority, a government ministry, or a corporate board of directors to organize the capital and manage the logistics. But Brian argues we are currently entering what he calls the abundance interregnum. It's a messy mess. middle period. The old king is dying, but the new law hasn't been sworn in yet. We are transitioning from
centralized scarcity to decentralized abundance. Let's talk about the historical shift in power. Brian writes that historically, empires controlled people by standing between them and scarce materials. Right. Joke points. Think about the Bronze Age empires guarding tin routes. The British Empire controlling coal and trade routes. Modern geopolitics fighting over oil or rare earth metals. If you control the hole in the ground, you control the people who need what comes out of it. The embargo is the ultimate weapon of the state. But programmable matter turns material scarcity into a temporary
software limit. If a neighborhood can build a quantum sensor or a medical membrane or a solar skin out of carbon stripped from the air and scrap metal pulled from the junkyard, the embargo loses its power. Because what are you embargoing? Dirt. Air. Exactly. As Brian bluntly puts it, you cannot centrally plan or embargo a million inverse design loops running in a million local rooms. He argues this will have a profound impact on both corporations and governments. Massive gigafactories and sprawling pharmaceutical campuses
will hollow out. They won't disappear entirely, but their role will shrink. The high -margin core of a business won't be the physical factory anymore. It will be the software, the unique AI weights, the proprietary design, and the trust they build with consumers. And for governments, the transition is equally stark. A state that rations steel, rations silicon chips, and rations medicine has immense leverage over its citizens. But when those goods assemble themselves from local atoms, the state must transition. Brian
says governments will have to rediscover a smaller, more fundamental job, keeping the peace and mediating disputes among people who no longer need them to eat or build. I have to admit, when I first read the section, my mind immediately went to like. Mad Max. Right. Total chaos. Yeah. If governments lose control of production, if the massive corporations hollow out, doesn't this mean anarchy? Doesn't the whole system just collapse into chaos? It's a very valid fear, and it's the first place most people's minds go. But Brian explicitly addresses this to reassure his readers. He states very
clearly that this does not mean anarchy, and it doesn't mean the courthouse burns down. Okay, good. It simply means the monopoly on necessary production expires. So the institutions survive the dynamic changes. Yes. Institutions will still survive, but they will survive because people freely choose to keep them around for organization and community, not out of coercion or the threat of starvation. It's a profound shift from compliance by necessity to cooperation by choice. But as these massive external structures change, as
the gigafactories hollow out and the supply chains collapse, what happens to the individual? If I'm not defined by my 9 -5 job at the massive centralized plant, Who am I? This is where Brian introduces a deeply philosophical element, drawing on the hero's journey framework by Joseph Campbell. He talks about the unbundling of jobs from identity. Right. For generations, civilization has commanded us, you are your job. You are your production quota. You are your salary. And when that command dissolves, when the machine can make anything
perfectly, people face a dark night of the soul. If the AI can inverse design the perfect battery, what is my purpose as an engineer? It's an existential crisis. But the solution Brian maps out in the 5 ,000 Days series is a return to craft. We won't just sit around in a hedonistic stupor. We will form voluntary guilds, mutual benefit circles, and operate household microfactories. We will treat AI and robots as tools under human direction, not as a new clergy that we worship. Exactly. We stop being cogs in a machine, we stop being mere employees, and we return to being artisans.
He writes this beautiful line in the piece, making things becomes optional as wage and available as joy. Wow. We build because we want to, because we want to infuse our humanity into the object, not because we have to hit a quota to pay for groceries. When you synthesize all of this, think about a student graduating right now in 2026. Oh, man. Yeah, they are stepping into a world where the career they were sold by their guidance counselor essentially no longer exists. The factory job, the standard engineering job, the logistics management job, it's all shifting beneath their
feet. It's all changing. But if they do the inner work, as Brian says, if they embrace this transition, they get something much better. They get a sense of purpose they can literally hold in their hands. They become the authors of their own physical world. It is a profound recentering of human potential. We are moving from an age of extraction to an age of pure creation. So let's recap this epic journey we've just been on today. We started with an AI at Oak Ridge National Laboratory looking through a scanning tunneling microscope and autonomously
nudging 37 molecules to build functional artificial graphene, spelling out a new era of engineering matter and proving that unsupervised atomic authorship is real. We explored the mechanics of that breakthrough, from the high -dimensional reinforcement learning brain to the YOLO computer vision eyes. We saw how that laboratory breakthrough merges with commercial tools like Atlant 3D's nanofabricators, moving us from prospectors discovering materials to authors inverse designing them. We explored the 25 -year arc toward local platforms, where
we will build our own batteries, medical materials, and quantum electronics from common dirt, scrap metal, and air. And we unpacked the massive societal shift that inevitably follows. The end of material scarcity, the decentralization of global production, the hollowing out of centralized power structures, and the beautiful return of the human artisan. It is the final lock on the age of abundance opening before our eyes. And I want to remind you, the listener, one more time. If you found these insights valuable today, please go support Brian Rommel's ongoing research. Yes, please do. Become a member at readmultiplex .com, buy
him a coffee, engage with the 5 ,000 Days series. We are navigating a totally new world, the pace of change is blinding, and voices like his are essential guides to understanding what happens next. As we close today, I want to leave you with a thought to mull over. Something to think about long after this deep dive ends. Let's hear it. If the cost of material goods drops to near zero, if our basic physical needs, energy, medicine, shelter, electronics, are met by a localized box in the garage running on sunlight and scrap. What becomes the new luxury? When anyone can
press a button and print a structurally perfect diamond or a flawless quantum sensor, perhaps the only true scarce resource left on Earth will be human attention. Or perhaps the ultimate luxury will be the completely unengineered, flawed, unpredictable beauty of nature. A wooden chair with a knot in it. A hand -thrown clay pot that isn't perfectly symmetrical. What will you choose to value when you can have absolutely anything? A brilliant, haunting question to end on. Thank you for joining us on this deep dive into the 5 ,000 Day Series. Keep questioning the future,
keep looking closely at the world around you, and we will see you next time. Welcome to the multiplex, where the future comes alive. Past, present, future tech, open up your eyes. Brian Romella leads the way with research deep and true. Technology and history, he connects them all for you. Brian Romella leads the way with research deep and true. Technology and history, he connects them all for you. Ancient sparks to modern code, he traces every thread, go explore
it. Thank you. for AI zero human companies building
from the start thermodynamic wages earning the key secrets we will read experiment in publishing research
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Every cell in your body has lived its whole life inside a magnetic field. The Earth is a dynamo, and it has been...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 3: From Steel to Hillbilly (Early Country). From steel to hillbilly: Frank Ferera,...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 2: The Knife and the Steel (Hawaiian). The knife of the blues slide guitar and the...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Somewhere in your house there is a computer you stopped thinking about. Maybe it is a 2017 laptop in a drawer, its...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 1 (Pilot): Blues. Fourteen original 78 RPM records from 1917 to 1925 that trace...