So Daniel sent us this one... he is looking at the future evolution of GSM. He wants to know what six G actually brings over five G, if there is a literal maximum limit to how much data we can fit over the air, and why the physical frequency and wavelength have such a massive impact on whether a signal can actually get through a concrete wall or not. This is one of those topics where the physics of the universe basically says, sorry, you can't have your cake and eat it too.
Herman Poppleberry here, and I have been diving into the white papers on this all morning. It is a fascinating intersection of pure electromagnetic theory and the very messy reality of urban construction. We are basically trying to fight the laws of physics to get high-speed internet into a basement.
Which, as anyone who has ever tried to send a text from a parking garage knows, is a losing battle most of the time. By the way, quick shout out to our invisible scriptwriter today—today's episode is actually powered by Google Gemini three Flash. It is helping us navigate the terahertz gaps. So, Herman, let's start with the immediate frustration. Why does my phone work perfectly on the sidewalk, but the second I walk behind a thick concrete pillar, it is like I have entered a lead-lined bunker?
It really comes down to the relationship between wavelength and the physical dimensions of the obstacles in the way. When we talk about five G or the upcoming six G, we are moving into much higher frequencies. High frequency means a very short wavelength. For context, old school four G LTE at seven hundred megahertz has a wavelength of about forty-three centimeters. That is roughly the size of a large pizza.
A pizza-sized wave. I can visualize that. It feels substantial, like it could actually push through some resistance.
Now, compare that to five G millimeter wave, which sits around twenty-four to thirty-nine gigahertz. Those wavelengths are between seven and ten millimeters. That is about the width of a pencil eraser. When a wave is forty centimeters long, it can practically ignore small obstacles. It is large enough to diffract, or bend, around corners and pass through materials without getting scattered by every little imperfection. But when your wave is the size of an eraser, every pebble in the concrete, every piece of steel rebar, and even raindrops become massive barriers.
So it is like trying to throw a basketball through a chain-link fence versus trying to throw a handful of sand. The basketball—the long wave—either hits or it doesn't, but the sand just gets scattered everywhere.
That is a decent way to think about it, although with radio, it is even more specific to the material's molecular structure. Concrete is a nightmare for high frequencies because it is porous and it holds moisture. Water is a polar molecule, meaning it loves to absorb electromagnetic energy, especially at certain resonant frequencies. There is actually a massive absorption peak around twenty gigahertz where water molecules just soak up the energy and turn it into tiny amounts of heat. So, when you try to send a high-frequency five G or six G signal through a damp concrete wall, the wall is literally eating your data and turning it into a microscopic amount of warmth.
Great, so my basement isn't getting internet, but it is technically getting point-zero-zero-zero-one degrees warmer. That is a comforting trade-off. But wait, if water is the enemy, what happens when it rains? Does a six G network just stop working the moment a storm rolls in?
Rain fade is a massive engineering hurdle. In the terahertz range, a heavy downpour can attenuate the signal by tens of decibels per kilometer. It’s not just water, either; even oxygen molecules have specific absorption bands. At sixty gigahertz, for example, there’s a huge spike where oxygen just gobbles up the signal. You could have a clear line of sight, but if the air is too "thick" with the wrong molecules, your range drops from kilometers to meters.
It’s wild that the very air we breathe is a barrier to the internet of the future. But before we get too deep into the six G future, I want to stick on this penetration thing. You mentioned rebar and conductivity. Why does metal specifically kill the signal?
This is due to something called the skin effect. When an electromagnetic wave hits a conductive material like the steel rebar inside reinforced concrete, it induces a small electrical current on the surface of that metal. At low frequencies, that current can penetrate deeper into the conductor, and some of the wave's energy can actually pass through or around the grid. But as the frequency goes up, the skin depth—the distance the wave can penetrate into the conductor—gets shallower and shallower. At the frequencies six G is targeting, which are in the terahertz range, the skin depth is practically zero. The metal reflects the signal like a mirror reflects light.
So we are building these hyper-connected cities out of materials that are essentially radio cages. It seems like a massive design flaw in modern civilization. We want gigabit speeds, but we also want buildings that don't fall down, and those two things are currently at odds.
It really is a fundamental tension. In the early days of GSM, back in the two G and three G eras, the network was designed for coverage. They used nine hundred megahertz because one tower could cover miles and miles, and the signal would happily float through your house, your office, and your car. But the trade-off was bandwidth. You can't fit a lot of data into a low-frequency wave because the "bins" for that data are physically large and happen less often per second. To get more data, you need more cycles per second—higher frequency.
Right, more vibrations equals more opportunities to encode a one or a zero. But those vibrations are so fragile they can't even handle a piece of drywall. So, that leads us to the big question Daniel asked: What is the actual limit? Are we going to eventually hit a wall where we just can't cram any more data into the air?
That is where we have to talk about the Shannon-Hartley theorem. This is the absolute mathematical speed limit for communication. It was named after Claude Shannon, the father of information theory. The formula is Capacity equals Bandwidth times the log of one plus the Signal-to-Noise ratio.
Okay, break that down for someone who didn't spend their morning reading IEEE journals.
Basically, there are only two knobs you can turn to get more speed. You can increase the bandwidth, which means using a wider slice of the radio spectrum—moving from a small one-lane road to a sixteen-lane highway. Or, you can increase the signal-to-noise ratio, which means making the signal much louder or the background "hiss" much quieter.
And since the world is getting noisier with more devices, we are mostly just building wider highways, right?
Precisely. That is why six G is looking at the terahertz range, which is between one hundred gigahertz and three terahertz. This is basically the "Wild West" of spectrum. It is completely wide open. We are talking about bandwidths that are hundreds of times wider than anything we have used before. Theoretically, six G could hit one terabit per second. For context, that is a hundred times faster than the peak speed of five G. You could download fifty high-definition movies in one second.
One terabit per second over the air. That sounds like science fiction. But if we already established that millimeter waves in five G struggle with rain and walls, terahertz waves must be even worse. If a pencil eraser wave is stopped by a wall, what does a sub-millimeter wave do?
It behaves almost exactly like light. At that frequency, the waves don't really "flow" around things anymore. They travel in straight, pencil-thin lines. If a person walks between you and the six G base station, the signal is gone. If a bird flies past, the signal drops. Even oxygen and water vapor in the atmosphere start to absorb the energy at those frequencies. It is incredibly volatile.
So if it is that fragile, how is it ever going to be useful? I don't want to have to stand perfectly still and hold my breath just to check my email. How does a network even maintain a connection if a single leaf blowing in the wind can break it?
That is the big shift in thinking for six G. We are moving away from the idea of "penetration" and toward the idea of "reflection." Instead of trying to blast a signal through a wall, six G is going to use something called Reconfigurable Intelligent Surfaces, or RIS. Imagine these as "smart wallpaper" or digital mirrors placed on the sides of buildings and inside rooms. These surfaces can actually detect where your device is and electronically tilt their reflection to bounce the signal around a corner and hit you directly.
So the entire city becomes a giant game of billiards with radio waves. The tower shoots at a wall, the wall bounces it to a lamp, the lamp bounces it to your glasses. But wait, does that mean the "mirror" has to move physically? Like a tiny motorized panel?
Not at all. These are meta-surfaces. They use an array of tiny antenna elements that can change their phase shift electronically. By adjusting the phase of each element, you can "steer" the reflected wave in any direction without moving a single physical part. It’s essentially a phased array antenna that acts as a mirror.
That is exactly the vision. It is called "Ubiquitous Intelligence." The network itself becomes aware of the physical environment. And because these waves are so high-frequency, they actually work like high-resolution radar. A six G network won't just send data; it will "see" the room. It could detect a person's posture, their breathing rate, or if they have fallen down, all without a single camera, just by analyzing how the terahertz waves bounce off their body.
It’s called Joint Communication and Sensing, or JCAS. In five G, we have beamforming, where the tower tries to point the energy at you. In six G, the tower and the environment work together to create a 3D map of the space. It can tell the difference between a human and a mannequin based on the microscopic movements of the chest while breathing.
That is both incredibly cool and slightly terrifying. It is like the walls have eyes, but the eyes are made of radio waves. Does this mean the network knows exactly where I am at all times, down to the millimeter?
Potentially, yes. The localization accuracy of six G is expected to be within centimeters. While that’s great for industrial robots or self-driving cars navigating a tight alleyway, it definitely raises some privacy questions. But from a technical standpoint, it is the only way to overcome the physics. If we can't go through the concrete, we have to go around it. This also means the "cell tower" as we know it is going to change. Instead of one big tower on a hill, we are going to have "In-X" subnetworks. Every room, every vehicle, every street light will basically have its own tiny access point.
It sounds like a lot of infrastructure. But let's go back to the "limit" question. Even with smart mirrors and millions of tiny towers, is there a point where the air itself is full?
There is a theoretical physics limit called the Bekenstein bound. It basically says there is a maximum amount of information that can be contained within a finite region of space with a finite amount of energy. Now, we are nowhere near that limit with radio waves—we are talking about orders of magnitude away—but it does mean that "infinite data" is physically impossible. The more practical limit is spectrum exhaustion. We are running out of the "good" frequencies that penetrate walls, which is why we are being forced into these difficult high-frequency bands.
It is like real estate. All the beachfront property—the low frequencies—was bought up decades ago by TV stations, the military, and early cell providers. Now we are trying to build skyscrapers on a swamp.
That is a great analogy. And the swamp—the terahertz range—requires much more complex engineering to stay stable. But there is another side to this. Some researchers think six G will be the end of the smartphone entirely. If you have one terabit speeds and sub-millisecond latency, you don't need a powerful computer in your pocket. You just need a pair of lightweight glasses that can stream high-definition three-D environments from a server nearby. All the processing happens in the "edge cloud," and the six G link is the invisible cord.
I have heard the "death of the phone" prediction for about ten years now, and yet here I am, still staring at a glowing rectangle. But I can see how six G makes that more plausible. If you can move that much data that fast, the device itself just becomes a screen and a battery.
But it brings us back to Daniel's question about the physics. Why does the actual frequency make such a difference? We talked about wavelength, but there is also the concept of dielectric loss. Materials like glass, wood, and concrete have a property called permittivity. At low frequencies, these materials are mostly transparent to radio waves. But as the frequency increases, the molecules in the material can't keep up with the rapidly oscillating electromagnetic field. They start to lag behind, which creates friction at a molecular level. That friction is what turns the radio wave into heat.
So it is quite literally a speed limit for the molecules in the wall. You are vibrating the field so fast that the wall itself says, "I can't do this," and just absorbs the energy.
That is a very high-level way of putting it, yes. And it is why five G and six G will always struggle with the "indoors" problem. Even if you have a massive six G tower across the street, if you have energy-efficient windows—which often have a thin metallic coating to reflect heat—your house is basically a Faraday cage. It will reflect the six G signal right back out into the street.
This explains why my signal drops the second I step into a modern office building but stays strong in an old wooden house. The wood is radio-transparent, but the "green" building is a data desert. Is there any way around that? Or are we just doomed to live in high-tech bunkers?
There's actually a fun fact about this: some newer skyscrapers are being built with "frequency-selective surfaces" or FSS. These are patterns etched into the window coatings that allow specific cell frequencies to pass through while still reflecting the infrared heat from the sun. It’s like a filter for your building. But it’s expensive, and it has to be tuned to the specific bands the carriers are using.
That sounds like a nightmare for city planning. "Sorry, we have to rip out your windows because you can't get six G." I can see the HOA meetings now. But let's talk about the practical side for listeners. If I'm looking at my phone right now and I see three bars of five G, what am I actually seeing? Is that a "good" frequency or a "fast" frequency?
This is one of the biggest marketing tricks in the industry. Those "bars" don't tell you the frequency. You could have five bars of a low-band seven hundred megahertz signal, which is great for a phone call but might be slow for data. Or you could have two bars of a millimeter-wave signal that is fifty times faster. The bars mostly measure signal strength, not quality or capacity.
So the bars are a lie. Or at least, they are a very incomplete truth.
They are a simplified metric for a very complex environment. If you want to know how your connection will actually perform, you have to look at which frequency band you are on. In the transition to six G, this will become even more polarized. You will have "coverage bands" that keep you connected and "data bands" that give you those crazy speeds, but they will flicker in and out as you move through a city.
It feels like we are trading reliability for raw power. I'd rather have a consistent one hundred megabits than a terabit that drops if I turn my head the wrong way. How does the network handle that handoff? If I’m moving at sixty miles per hour in a car, how does a "pencil-thin" beam stay locked on me?
That’s one of the biggest challenges of six G—mobility management. To keep that beam locked on a moving target, the network has to predict your movement. This is where the AI comes in. The base station isn't just reacting to where you are; it’s predicting where you’ll be in the next ten milliseconds based on your current trajectory. If it misses, the connection drops. It’s like trying to keep a laser pointer on a fly.
I appreciate the effort, physics, I really do. But it sounds like the network is playing four-D chess just to make sure my Instagram reel doesn't buffer.
It is a massive coordination problem. Think about the connection density. Five G is aiming for a million devices per square kilometer. Six G is looking at ten million. That is not just phones; that is every sensor in the road, every smart lightbulb, every wearable device, and even "zero-energy" sensors that harvest power from the radio waves themselves.
Wait, harvest power? Like, the radio wave is the battery?
Yes! That is another six G goal. Ambient backscatter communication. You could have a tiny sensor on a package or a bridge that doesn't have a battery at all. It just sits there, and when the six G signal hits it, it uses a tiny bit of that electromagnetic energy to wake up, send a bit of data, and go back to sleep.
Okay, that is a game changer for IoT. No more changing batteries in a thousand different sensors. But that also means we are pumping even more electromagnetic energy into the air. Is there a health limit to how much of this stuff we can have around us?
There are very strict international standards on SAR—Specific Absorption Rate. Because these high frequencies don't penetrate the skin—they are absorbed in the very top layers—the concern isn't about internal organs like it might be with lower frequencies. It is more about surface heating. But the power levels we are talking about are so low that the heating is negligible. You get more electromagnetic radiation from standing in the sun for ten seconds than you do from a lifetime of six G.
Good to know. I'll keep my hat on and my six G glasses on. So, to wrap this part up, the reason concrete blocks us is that we are trying to use waves that are too small and too fast for the material to handle. It is like trying to drive a Ferrari through a thicket of trees. You might have the speed, but you don't have the clearance.
And the future is about building roads—or mirrors—around those trees. We are moving toward a world where the distinction between "online" and "offline" disappears because the network is woven into the physical fabric of our buildings.
It is a fascinating roadmap. I want to shift gears a bit and talk about the takeaways. What should people actually do with this info? Because it is one thing to know about terahertz gaps, but it is another to apply it when you are buying your next house or phone.
One big thing is understanding the trade-off in your home setup. If you are moving into a new apartment or building a house, you need to think about internal wiring. People think mesh Wi-Fi is the answer to everything, but at the frequencies we are moving toward, even Wi-Fi six E and seven are starting to hit these same penetration issues. If you have the chance, run fiber or high-quality ethernet to every room. Don't rely on the signal getting through the walls, because the walls are only getting "thicker" from a radio perspective.
That is such a practical point. We are becoming more wireless, but to make that work, we actually need more wires inside the walls to feed the local access points. It is a paradox.
It really is. The more wireless the world becomes, the more glass and copper we need in the ground. Also, for smart home planning, keep in mind that "long-range" and "high-speed" are usually mutually exclusive. If you want a smart lock on your front gate that is fifty yards away, you want a device that uses a low-frequency protocol like LoRa or old-school Zigbee, not something trying to run on a high-frequency band.
Right, use the "pizza waves" for the gate and the "eraser waves" for the living room.
And another thing for the tech-savvy: start looking at signal mapping tools. There are apps that let you see exactly which frequency band your phone is connected to. If you are getting poor speeds despite having "full bars," check the band. If you are on a high-frequency band with a weak signal, moving three feet closer to a window might jump your speed from ten megabits to five hundred.
I've actually done that. It feels like you are doing a weird dance with the invisible spirits of the air, but it works. You find that one "hot spot" near the curtain and suddenly you are in the future.
You are literally finding the path of least resistance for those tiny waves. And as we move to six G, those hot spots will become more defined. You might have a "high-speed chair" in your living room where the RIS mirror is aimed, and if you move to the couch, you drop back to five G.
I can see the real estate listings now: "Three-bedroom, two-bath, excellent terahertz coverage in the master suite." It is a whole new layer of value. So, looking ahead, is there anything that could disrupt this whole six G terahertz roadmap? Like, could we find a way to make concrete transparent?
There is actually research into "radio-transparent concrete." They use different types of aggregates and binders that don't hold as much moisture and don't have the same dielectric loss. But the problem is cost and structural integrity. You have to balance "can it hold up a bridge" with "can I watch YouTube through it." Most engineers are going to choose the bridge.
Probably a wise choice. I prefer my bridges not to collapse, even if it means I can't stream in the middle of them. Does that mean we’ll see a split in construction styles? Like, "tech-friendly" buildings versus "old-school" buildings?
We are already seeing it. Some high-end office spaces are being marketed as "RF-transparent." They use specialized drywall and internal glass partitions that are designed not to scatter the gigahertz signals. It’s basically interior design for the electromagnetic spectrum.
Another disruption could be satellite. If Starlink and its competitors can get latency low enough and density high enough, we might bypass the "ground-based tower" problem for a lot of things. But even satellites have to deal with the atmosphere and building penetration. You still can't get a satellite signal inside a basement.
Satellites are just very high towers. They still face the same physics. In fact, terahertz from space is almost impossible because of the atmospheric absorption we talked about. Space-to-ground links usually stay in the lower gigahertz range specifically to avoid being eaten by the clouds.
Physics remains the ultimate boss. You can't bribe it, you can't negotiate with it, you just have to work around it.
And that is the beauty of it. The history of GSM is really a history of humans finding increasingly clever ways to cheat the limitations of the electromagnetic spectrum. From two G's simple digital pulses to six G's AI-controlled smart mirrors, we are just getting better at the game.
It is a game of inches—or millimeters, in this case. I think we have covered a lot of ground here. We started with why your phone dies in a garage and ended with smart wallpaper that can sense your heartbeat. It is a wild trajectory.
It really is. And big thanks to Daniel for the prompt. Every time we dig into this, I realize how much of our modern world relies on these invisible, fragile waves that we just take for granted.
It is the invisible architecture of the twenty-first century. Well, I think that is a wrap on the physics of the "radio cage." Thanks as always to our producer Hilbert Flumingtop for keeping the gears turning behind the scenes.
And a huge thank you to Modal for sponsoring the show. They provide the serverless GPU credits that power our entire pipeline, including the generation of this very script.
This has been My Weird Prompts. If you enjoyed our deep dive into the terahertz future, leave us a review on Apple Podcasts or Spotify. It actually helps more than you'd think in getting the show in front of new people.
We will see you in the next one. Hopefully, with a clear line of sight.
Stay connected, everyone. Goodbye.
Goodbye.
