Who Owns the Memory of the Earth?

I went looking through my email for the first mention of Planet and found it in a Fortune newsletter from July 2015. It was buried among funding rounds, acquisitions, and the usual movement of money through technology companies. Planet Labs had acquired RapidEye and some other geospatial assets. The email mentioned that the company had recently raised $118 million. At the time, it would have been easy to read this as another startup growing through capital and acquisition, then move on to the next item. I probably did exactly that.

A few months later, another email said the acquisition had been completed. In 2017, Planet appeared again when Google sold it the Terra Bella satellite business. These early messages were about the company itself: how much money it had raised, what it had bought, and how quickly it was expanding. Then the references in my inbox began to change. Planet appeared less often in headlines and more often beneath images. A military facility in North Korea. Chinese construction near a disputed border. Ships gathering at sea. Damage after an earthquake. Roads cutting into forests. The small line underneath would say that the image came from Planet Labs.

Somewhere during those years, Planet stopped being only a company I was reading about and became part of how other stories were proven.

I find this shift interesting because it is often how infrastructure enters our lives. At first, the company is visible. We talk about the founders, the funding, and whether the idea will work. Later, the company recedes and the system becomes ordinary. It sits underneath journalism, finance, science, environmental monitoring, and war. We start relying on it without thinking much about where it came from or who controls it.

Planet has placed hundreds of small satellites around the Earth. They repeatedly photograph the land below and add those images to a growing archive. The company now describes its work as observing change at the daily pace of the world itself. For many places, there are already years of images showing how the same patch of ground has changed over time.

I had thought about satellite images mostly as photographs from very far away. Looking through these emails made me realise that Planet is building something larger. It is building a memory for the physical world.

For most of history, much of what happened to the Earth disappeared without a clear record. A forest could be cleared slowly enough that no single moment looked dramatic. A mine could expand behind a fence. A river could be diverted. A government could construct something in an isolated region and deny what it was doing. The people nearby might understand the change very clearly, but their knowledge could remain local, oral, and easy for more powerful institutions to challenge.

Now we can return to the same place again and again. We can look at what was there last year, last month, or last week. We can see when a road appeared, when trees began to vanish, when new buildings emerged, or when military equipment moved into an area. The ground is beginning to have something like a version history.

This becomes even more powerful once AI starts reading the archive. There are already too many images for people to examine one by one. A machine can compare them and search for particular kinds of change. It can notice that vegetation is becoming less healthy, that methane may be leaking from an industrial site, that a logging road has appeared inside a protected forest, or that ships are gathering somewhere unusual. The useful output is increasingly an alert rather than a picture.

The system begins to tell us that something happened and that we should pay attention.

There is real value in this. A small environmental organisation cannot visit every part of a forest every day. A community may know that land is being damaged and still struggle to make anyone listen. Satellite evidence can help them show where the damage happened and how quickly it spread. Journalists can investigate places they cannot safely enter. Researchers can understand changes happening across areas that would be impossible to study only from the ground.

I can understand why Will Marshall, Planet's co-founder, speaks about the technology with such hope. He talks about making global change visible, accessible, and actionable. The idea is that seeing the Earth more clearly could help us become more responsible toward it. We have created many systems whose consequences are spread across such large distances that most of us never see them. The product we buy in one country may be connected to a forest disappearing in another. A company reports its environmental performance through numbers, while the land itself tells a different story. Satellite imagery can bring some of those distant consequences closer.

It can make abstraction physical again.

I also know that an image does not carry its own moral purpose. The same satellite that sees illegal logging can see military movement. The same archive that helps document environmental destruction can support border enforcement, commodity trading, insurance pricing, and surveillance. A camera in orbit does not know whether someone is looking in order to protect a place, profit from it, or control the people living there.

That part depends on who gets access and what power they already have.

This is where the idea of a planetary memory starts to feel less simple. Planet is recording a shared world, but the record belongs to a private company. The land being photographed may belong to a community, a farmer, a state, or nobody in any meaningful sense. The image still enters a commercial system. Access depends on subscriptions, contracts, technical capacity, and the value a customer can extract from the data.

The people living inside an image may never be allowed to see it.

Planet increasingly presents itself as a platform that can help organisations look broader, backward, closer, and deeper. I understand the appeal of that language. A system like this turns a huge amount of complicated information into something people can search, compare, and act upon. It could help us understand crops, forests, fires, coastlines, cities, and climate risks.

It also reveals who these systems are usually designed for. The most capable interfaces are often built for people who already have the resources and authority to make decisions. They receive a clearer view of the world so they can move faster inside it. A planetary intelligence platform could become a tool for planetary care. It could also become a control panel used from offices far away from the communities whose lives appear in the data.

The surface of the Earth is gradually becoming machine-readable. A new roof can suggest construction activity. The colour of a field can reveal crop stress. The number of vehicles outside a factory can become an economic signal. A road inside a forest can trigger an environmental warning. Ships moving through a harbour can affect commodity markets before anyone nearby understands why prices have changed.

The same pixels can tell many different stories. The story that becomes important usually belongs to the person with the ability to ask the question.

This is why I do not think the main issue is whether satellite monitoring is good or bad. The technology is already part of both care and control. Military imagery can expose war crimes. Environmental data can be used to remove Indigenous people from land in the name of conservation. Agricultural monitoring can support farmers while giving traders information those farmers do not have. Climate-risk models can help a city prepare for floods while making poorer neighbourhoods impossible to insure.

Seeing changes the balance of power, but it does not always move power in the same direction.

The questions that stay with me are practical ones. Can a community access the historical images of the place where it lives? Can people understand how an algorithm has interpreted changes to their land? Can they challenge that interpretation? Which parts of the archive should be treated as public knowledge? What happens when satellite imagery is combined with location data, financial data, mobile-phone records, and other forms of surveillance?

We are building the memory before we have agreed on the rights around it.

Looking through the emails, I can see Planet becoming part of how reality is verified. It begins as a company mentioned in a funding newsletter. Eleven years later, its images sit underneath stories about forests, borders, wars, disasters, agriculture, and climate. The company becomes less visible while its view of the world becomes more important.

There is something reassuring in the idea that the physical world is becoming harder to rewrite after the fact. A government can deny that construction took place. A company can claim that no forest was destroyed. An army can dispute when its equipment arrived. The archive remains. Someone can go back and look.

The Earth begins to remember what happened to it.

I want that memory to exist. Too much damage has disappeared because the people responsible had more authority than the people who witnessed it. Too many communities have been asked to prove what they already knew from living inside a place. A persistent visual record could give them evidence that survives denial.

I also do not want the memory of the Earth to become another system that gathers knowledge from everyone and concentrates its value somewhere else.

This tension will not be resolved by the satellites themselves. It will be shaped by access, ownership, policy, design, and the choices made by the organisations using the data. It matters whether these systems are built only for governments and large companies, or whether ordinary people can use them to understand and defend the places they care about.

There are already partial examples of a different direction. Planet makes some imagery available to researchers through programmes such as the European Space Agency's Earthnet. Environmental organisations use high-frequency images to detect forest change earlier. Journalists and open-source investigators use commercial imagery to make distant events visible to the public. These arrangements are valuable, but they still feel like exceptions negotiated around a system whose centre remains commercial.

Perhaps the question is not whether the whole archive should be free. Running satellites, processing images, and maintaining an enormous historical dataset costs real money. The harder question is what kind of public rights should exist inside privately operated planetary infrastructure. A community may not need unlimited access to every image of the world. It may still deserve meaningful access to the history of the land where it lives, especially when that history is being used to make decisions about its future.

Planet is already photographing the Earth every day. AI is already learning how to read the images. The archive will keep growing whether most of us pay attention or not.

We used to photograph places because we wanted to remember them. Now we are building a system that photographs change itself. It can show us what is happening to our shared home with a consistency that no human witness could maintain.

The question is who will be allowed to look through that memory, whose interpretation will matter, and whether seeing more clearly will help us take responsibility for what we find.

Connected reading

• Sixty Bets Against Forgetting
• When a Website Loses Its Memory
• The World Runs on Companies You Have Never Heard Of
• The AI Future Is Becoming a Land Use Problem

AI authorship note

This essay began with Kris Haamer searching eleven years of his email for every reference to Planet Labs and noticing how the company moved from funding news into the image credits beneath stories about forests, borders, wars, and disasters. AI supported email retrieval, source checking, drafting, editing, SVG illustration code, and implementation. Kris directed the essay toward the idea that Planet is building a memory for the physical world, repeatedly rejected versions that felt too structured or distant, and asked for the final piece to remain personal, searching, and built from sustained paragraphs.