Two Revolutions That Are Often Confused

Part 4 of 30 · Series date: September 8, 2026

One spacecraft exports answers. Another exports electricity. Confusing the two hides the engineering that matters.

Imagine two satellites leaving the same rocket. Both unfold solar panels. Both point toward the Sun. Both communicate with Earth. A casual observer might call them two examples of the same technology.

But one sells computation, while the other is designed to sell electricity. Their resemblance hides a major difference in what customers receive—and what must work before either customer can be served.

One exports answers

In the first satellite, solar cells supply electricity to onboard equipment. A processor examines data or runs an application. A communications terminal returns the result.

The useful export is information. A completed scientific calculation might be summarized in a small file even if producing it required extensive computation. This separation between the size of a result and the work needed to obtain it is one reason orbital computing attracts interest.

It is not universal. Some applications require enormous inputs, generate enormous outputs, or exchange information continuously with other machines. The networking requirement must be calculated for the actual application.

The other exports energy

The second satellite converts its electricity into a transmitted beam. A receiver converts part of that beam back into electricity. JAXA's space-solar research explicitly considers microwave and laser delivery architectures. JAXA: Space Solar Power Systems.

Here, the transmitted energy is the product. There is no clever summary file that substitutes for delivering the promised kilowatt-hours. A household cannot cook dinner using a message saying that enough energy was generated elsewhere.

Conversion efficiency, beam capture, atmospheric conditions, receiver availability, and grid connection therefore sit directly in the business model.

Follow an imaginary hundred units

Consider a deliberately simplified power-beaming example. Suppose a solar system supplies 100 units of electricity to a transmitter. Assume the transmitter converts 80 percent into the beam, the receiver captures 80 percent of that, and conversion plus delivery retains another 80 percent.

The final delivery is 100 × 0.8 × 0.8 × 0.8, or 51.2 units. These assumed efficiencies are teaching numbers, not a forecast for any proposed plant. Solar-cell losses occurred before the starting point and are not included.

The lesson is that efficiencies multiply. A series of individually good components can still lose a substantial share of the starting energy. Some losses become heat, and energy missing the receiver must also be accounted for.

An onboard computer avoids the Earth-directed power chain, although it still has internal electrical losses and communications costs. That is an architectural difference, not proof of profitability.

Shared tools, separate customers

The industries could benefit from common advances in lightweight arrays, launch packaging, reliable deployment, power electronics, and robotic maintenance. A manufacturer might supply components to both.

Customers would judge them differently. A compute buyer wants useful results at a promised price, speed, and quality. An electricity buyer wants dependable energy with clear delivery terms. A utility may need support during specific hours; an AI customer may instead accept a job completed overnight.

Even the unit of success differs. A computing operator might track cost per verified task. A power operator might track delivered cost per megawatt-hour and availability during contracted periods.

Combining the two under a vague promise of “abundant space energy” makes these practical questions harder to see.

A third possibility: power for other spacecraft

Power beaming does not have to end on Earth. A generator might serve another satellite or a lunar installation. Such customers could have different alternatives and different willingness to pay.

Imagine a robotic explorer that needs power for a short operation where a cable is inconvenient. A beam could be worth investigating even if the same technology cannot compete with ordinary grid electricity. Conversely, supplying a nearby computer through a wire may be simpler than building a separate power spacecraft and transmitter.

These are design options, not deployed services assumed by this article. Distance alone does not choose among them. Movement, pointing, duty cycle, safety, and backup needs also enter the decision.

Three products with three receipts

A simple comparison keeps the accounting straight.

The distinction is not bureaucratic. It identifies what the supplier must actually deliver. A computing customer can receive an excellent answer even if the machine needed more energy than expected; the supplier bears the cost unless the contract says otherwise. A power customer cannot receive excellent electricity in place of missing electricity. Quantity and timing are part of the product.

An attractive diagram can hide a double count

Imagine a proposal with a large solar array, an onboard computer, and a transmitter aimed at Earth. Its presentation promises enough power to run the computer and also advertises the array's entire output as electricity for sale.

Unless the claims refer to different operating periods, the same energy has been counted twice. Electricity used by the computer is not simultaneously available to the transmitter.

The operator could divide supply between the two uses or switch between them. It might compute when Earth-directed delivery is unavailable and export power when a receiving site pays more. That is a possible operating strategy, but it adds scheduling, equipment, and commercial obligations.

Waste heat does not restore the missing electricity. Recovering work from heat would require another system and a suitable temperature difference, with its own losses. A proposal cannot turn an accounting mistake into abundance merely by labeling waste heat a second energy source.

Separate energy ledgers would show generation, essential spacecraft consumption, computing consumption, transmission input, and losses for each operating period. This is the level of clarity a hybrid system needs before its apparent versatility can be trusted.

A successful technology transfer can be smaller than the vision

The two industries do not have to merge to help one another. A power-satellite team might develop a deployable structure that also supports a computing satellite's radiator. A computing operator might develop reliable remote fault diagnosis that helps maintain a power platform.

In these illustrative cases, the transferred product is a component or method—not proof of an integrated orbital economy. That smaller success could still be valuable.

It also offers a healthier way to read investment announcements. An organization may have a sound market for a component while its largest customer vision remains uncertain. Conversely, a persuasive market for electricity does not establish that a particular supplier can manufacture the required hardware.

Keeping the levels separate protects both enthusiasm and judgment. We can celebrate a real advance without requiring it to carry a claim larger than the evidence supports.

The next time a headline combines solar power and artificial intelligence in orbit, ask where the electricity is used and what leaves the spacecraft. Those two questions cut through much of the confusion before any complicated calculation begins.

What would prove this?

Ask every proposal to finish one sentence: “Our customer pays us to deliver ___.” Then identify every link between sunlight and that deliverable.

For computing, request a workload demonstration with complete energy and network accounting. For power, request measured energy at the receiver, not just transmitter output. For a hybrid, require separate accounts so one promising component does not hide another's poor economics.

The two revolutions may eventually strengthen each other. Keeping them distinct now is how we can recognize real progress in either one.

Previous article · Next article