Have you ever wanted to capture the entire energy output of the sun?

Well have I got the gadget for you! It’s a theoretical megastructure  called a Dyson Sphere, and it’ll encompass your entire home star.

If you’ve ever heard of the Kardashev Scale, the Dyson Sphere might sound familiar. Type II civilizations are  defined as those that can harness energy on the scale  of that emitted by a star. For example, by using a Dyson Sphere.

Although, as you might imagine, building a bubble around the  Sun is easier said than done. I’m still waiting for them to repair the pothole down the road from my house! To build something on the scale of a Dyson Sphere, we’d need to solve problems  like planetary-scale mining and complex gravitational dynamics, as well as ultra-efficient energy capture,  storage, and transmission.

While we definitely can’t start  building one of these today, we do already have some technology that could someday evolve into  Dyson-Sphere-construction-tools. So let’s take a journey into  the distant future and make a shopping list for our DIY  Solar System renovation project. [♪INTRO] The concept of a Dyson Sphere goes way back to a 1937 science fiction novel. But the first non-fiction consideration of energy collecting spheres around stars was a paper from 1960 by a physicist  named, you guessed it, Dyson.

No, not the vacuum guy, different Dyson. That paper from 1960 was a thought experiment about the tremendous amount of energy that an extraterrestrial civilization  would need to consume in order for us Earthlings to detect  it from the outside looking in. Some have interpreted it as a work of satire poking fun at the search  for extraterrestrial life.

And later in his life, Freeman Dyson himself went back and forth on the concept. Sometimes he referred to it as a joke, and other times stood by his proposal. Nevertheless, there have been multiple searches to see if other civilizations have built a Dyson Sphere anywhere near  our galactic neighbourhood.

So far, the answer is no. But  there was a false alarm in 2015, set off by dust around a star  commonly known as Tabby’s Star, located about 1,500 light years away. And while no one on Earth is currently planning to build a Dyson Sphere, we’ve collectively taken some  baby steps in that direction.

Like in 2019, China announced plans to build a solar power station  orbiting the Earth by 2035. And tech companies have seriously considered building AI servers in space. In orbit, solar panels can access sunlight 24/7, but dissipating heat could be  tough and repairs would be costly.

Even so, the first small scale test of a satellite-based AI server  launched in November 2025. Now there are also a whole bunch of ethical questions that come along with a civilization choosing to harness so much energy from their sun. It’s not a given that humanity would even want to go down that power hungry path.

But we can learn a lot by thinking about how we might build something so monumental and all the challenges we’d face along the way. So for this video, we’re going to  set those existential questions aside and figure out what it would take  to actually build the dang thing. Let’s get one thing out of the way, a solid Dyson sphere is essentially impossible  for quite a few reasons.

Even Dyson himself thought so. First of all, there’s no way to anchor it so that the Sun stays in the center. You can’t use gravity, because  inside an enclosed sphere, the net gravitational force is zero.

The pull from one side is  cancelled out by the other. The sphere exerts no net force on the Sun and the Sun exerts no net force on the sphere. So any other forces on the sphere  could bump it right into the Sun.

And there are some unexpected  forces out there in space, you know comets and asteroids and whatnot. It would also need to be  perfectly spherical because any imperfection could cause a gravitational imbalance which could also make it drift into the Sun. Plus you would need a massive amount of incredibly strong material that could withstand the brutal gravitational forces  on each panel of the sphere.

Yes I know we said there’s no net gravitational force from the Sun on a full sphere, but think of the individual panels that make up the sphere. Each panel is pulled inward by  gravity, squeezing it against its neighbouring panels, which all  need to bear this tense load. Similar to how every stone  in an arch squeezes against its neighbors to stand upright  against Earth’s gravity.

The panels would need to be made of material that are far stronger than  anything we know of today, and it’s not even clear we  would have enough stuff in our Solar System to construct  a continuous sphere that big. But there is a slightly more  practical version: a Dyson Swarm. Instead of a solid structure,  think of a collection of satellites all orbiting independently.

The big advantage here is you could build it incrementally and start  generating power right away. There’s nothing stopping  us from building just a few swarm satellites to test how they work before we commit to a Solar System–scale project. Of course that doesn’t mean it would be easy.

The worst-case scenario is a  collision between satellites that sets off a chain reaction  and brings the whole chain down. We could try to carefully  choreograph an orbit pattern to avoid collisions, but long-term, a swarm will require active upkeep and maintenance. Luckily, we already know how to do this.

Carefully choreographing and  repositioning satellites is something we do today for the 17,000 satellites we have in orbit as of early 2026. Although we might not need to worry about that if we built a Dyson Bubble instead. This is a similar idea to a swarm  with many independent structures, but each of the structures would  be tethered in a (relatively) stationary position, rather than  circulating in a traditional orbit.

Since orbits rely on angular momentum to balance the inward force of gravity, the physics of a Dyson Bubble needs to work a little differently. Photons of light exert a very small  amount of force when they strike a surface, so the components of a  Dyson Bubble could use solar sails to balance photon pressure with the  gravitational pull from the Sun. The photons pinging on the  inside of the sail could push it outward enough to counteract  the inward pull of gravity.

But the catch is these Dyson Bubble components would need to be ultra light. But we have already built solar sails like LightSail 2 which flew for over three years. So a Dyson Bubble isn’t out of the question.

Where we put our Dyson Swarm  is the next crucial question. And we’ll answer that question about our power-producing megastructure right after we power our production with this quick ad. Since you watch SciShow, I know you’re the kind of person who loves learning outside of a classroom.

And that’s what makes this video’s sponsor, Brilliant, so great too. Brilliant is an online learning platform made for everyone from age 10 to 110. Those courses are crafted by world-class teachers from MIT, Harvard, and Stanford, and they’re designed to build  your confidence in coding.

Take their course on Algorithmic  Thinking for example. In that course, you get to  solve over 200 problems. It’s far from falling asleep in a lecture hall!

To learn for free on Brilliant for a full 30 days, go to brilliant.org/scishow,  scan the QR code onscreen, or click on the link in the description. Brilliant’s also given our  viewers 20% off an annual Premium subscription, which gives you unlimited daily access to everything on Brilliant. The closer a satellite is to the Sun, the more power it can capture, requiring fewer satellites overall.

On the other hand, the most  efficient plan would be to mine the materials, construct the components, and launch the swarm satellites  all from the same place. So to strike the perfect balance, we need a planet with the perfect résumé. It needs to be made of the right stuff to mine, have an environment we could build things in, and have low gravity for  low-effort satellite launches.

Dyson had his eye on Jupiter for this role. But the gas giants don’t have the right elements, or a solid surface to build things  on, plus they’re far away from the Sun, and we would need to overcome  their high gravitational pull. So in 2022 a paper instead ran the numbers for the rocky, terrestrial  planets of the solar system.

It’s likely Earth would be out of the running for sentimental reasons. It’s the only place we know we can survive without life support. Kind of important.

But let’s say for a moment that we did want to build a swarm at Earth’s orbit. If we want our swarm to produce  power at the same rate we’ve been consuming it, then to make  all those satellites we would need to scrape 2.9 meters of material  off of all the land on Earth. Ouch.

On top of that, Earth has the highest  gravity of the terrestrial planets. It would require some heavy lifting to get each and every satellite off the ground. Looking at our neighbours, Venus’  gravity is close to Earth’s.

But with surface temperatures  of 465 degrees Celsius, it’d be a tough spot to put a manufacturing plant. Plus the probes we send to  Venus have a tendency of getting crushed by the  enormous atmospheric pressure. Mercury looks kind of promising.

It’s probably the most discussed source of materials for a Dyson Swarm. It’s close to the Sun, and small, so  it would be easy to launch things. It’s also made of the right stuff, being 70% iron.

A lot of it is in the planet’s  core, though, so it would take some digging to get to. Okay, by  “some digging” I do mean we would essentially need to  dismantle the entire planet. And I guess one side of Mercury is 449 degrees Celsius while the other is -170!

So you can’t have humans doing the work, and none of our current robotic technology could survive such extremes either. Maybe not so promising, after all. That leaves Mars.

Its surface temperatures range from  20 degrees Celsius down to -150. Chilly but feasible. And over there we kept the Opportunity rover running for nearly 15 years!

Mars’ crust is 14% iron,  so it’s realistic to mine, and its gravitational strength  is just under 40% of Earth's. So it makes sense that  particular study picked Mars as the winning candidate for Dyson Swarm production. Oh and we can’t forget, those  satellites need to capture some power.

One obvious way to do this is  with photovoltaic solar panels. They’re what we currently use  to power most of our satellites. Our current versions are unfortunately  expensive and break easily.

The tech is improving fast, but  we have a long way to go until photovoltaic solar panels are durable  enough to last for generations. It would be a hassle to repair panels on swarm satellites millions  of kilometers from Earth. Another possibility is concentrated solar power.

This would just mean putting  a bunch of big mirrors on the satellites and bouncing  sunlight toward collection stations. These could either be in space  or on a planet like Mars. When light hits a collection station, it would either hit a highly efficient solar panel or be absorbed as heat before  being converted to electricity.

We already do something like this  on Earth at thermal solar plants. Massive arrays of mirrors focus the Sun's rays on salts that melt at high temperatures. The energy is then stored as  thermal heat in the melted salt.

But for any of these options,  you’d additionally need to get the energy from the production  point out to where people need it. You think untangling Christmas lights is bad? Try running power cords to satellites!

Wireless transmission  through microwaves is one way to do this and we’ve even already  demonstrated it’s possible. In 2023, for the first time,  researchers beamed a small amount of power from a satellite to the surface  of the Earth through microwaves. Rome wasn’t built in a day, and a  Dyson Sphere wouldn’t be either.

We would need to build gradually,  iterating along the way. Construct and launch a couple sections, learn some valuable tech lessons from it, and use the energy that those sections  collect to build the next batch. Unfortunately it would take a LOT of rocket fuel to launch all those components into space.

So an electromagnetic launch system, essentially a giant railgun,  would likely be more practical. While we haven’t yet launched  any satellites from Earth using one of these, there are groups  already working on similar ideas. In 2021, a company completed their first suborbital test flight launched from a centrifuge.

This works by spinning a spacecraft very fast before launching it vertically. The mechanism is a little  different from a railgun, but it’s the same basic idea of  using something besides rocket fuel to give the craft enough  kinetic energy to reach orbit. Or that same 2022 study about the terrestrial planets proposed an alternative plan for building a swarm with a more easily achievable goal.

While satellites orbiting the  sun are certainly possible, their plan suggests they  should orbit Mars instead. It wouldn’t look like a traditional Dyson swarm, but it would be able to  more easily reflect light to planetary base stations by  keeping the satellites close by. They estimate it would take about 50 years of building to make a swarm that could generate power at the same rate as  the Earth consumed in 2019.

So it would be a slow start, and that plan would require 5.5 billion satellites  in orbit around Mars. But it’s not the upper limit. If you saturated the space around  Mars you could capture more than eight times as much power  as the Earth used in 2019.

You could charge so many Nintendo  3DS with all of that power. Or I guess, you know, like, build  some massive particle colliders or improve interplanetary travel. You know, there are options.

There’s no doubt we have a long way to go before any of this becomes reality. If it ever does. But even within our current  understanding of physics and engineering, we already have some tools that could evolve to Dyson Swarm–level tech.

Of course, the ethics and economics  of building one to consume that much energy are a whole  other complicated conversation. But hey, all we promised here was a project plan! Now, does anyone know a  project manager with experience building multi-generational,  theoretical megastructures?

Anyone? [♪OUTRO]