We Want to Build a Company That Does Not Turn People into Raw Material

1. How the Original Purpose Can Be Lost

We (Mona and I) have built companies, organizations, and social projects in our lives. Most began with an idea that made sense. With something we wanted to change, improve, or make possible in the first place.

Then came structures, costs, expectations, processes, and growth. All necessary. At some point, the question sometimes changes.

At the beginning, the question is: What impact do we want to have? Later, it becomes: How do we keep the system running?

This rarely happens with a great bang. No one gets up in the morning and decides to betray the original purpose. It happens through small, reasonable-sounding decisions. A little more revenue. Another metric. A feature intended to bring users back more often. A little more context because the result will then be better. A claim that goes further than the product can already support.

Every individual decision can be explained. We humans really are very good at that.

And at some point, you have built a company that still talks about its original values but has long since become economically dependent on something else.

That is exactly what we want to prevent at FlameP.

2. People Can Become Raw Material in Different Ways

When we think of raw materials, we think of oil, wood, water, or metals. Something is extracted, processed, and commercially exploited.

In the digital economy, people have already assumed this role far too often. We have all become raw material ourselves. Unless we are sitting in a cave in the Himalayas meditating all day.

Our data are processed. Our attention is sold. Our time is converted into time spent. Our uncertainty generates further use. Our language, decisions, and corrections improve systems that do not belong to us.

With artificial intelligence, this relationship becomes even closer.

A traditional platform may know which posts someone clicks, how long someone watches a video, and which product that person buys. An AI comes much closer to a person’s thinking. We tell it what we are working on, which decision concerns us, what we do not understand, and where we have doubts. We upload documents, describe relationships, and correct its answers until they match our language.

That can be helpful.

But it can also create increasingly valuable personal context while the person can barely trace who owns that context, where it is processed, and how dependent they have become on a single provider.

That is why the General Data Protection Regulation does not limit data protection to formal consent. It also establishes purpose limitation, data minimization, and data protection by design. Personal data should not simply be used because they are technically available, but only to the extent necessary for a clear purpose. [1]

Acting in a legally sound manner alone is not enough for us.

It is at least a start to ask whether data are processed lawfully. The relevant question is what kind of relationship a company builds with the people on whom its economic success depends.

3. We Do Not Believe Technology Is Evil

We are not opposed to technology. Otherwise FlameP would be a rather contradictory undertaking.

Artificial intelligence can take work off people’s hands, make knowledge more accessible, lower language barriers, and create opportunities that only a few years ago were reserved for large companies or highly specialized experts.

That is why the United Nations Human Development Report 2025 does not ask whether AI is fundamentally good or bad. It describes AI as a matter of human choice. The decisive question, it says, is whether AI expands people’s capabilities and agency. [2]

Technology has no principles. But products have rules. Business models have incentives. Metrics change decisions. And systems can push people in a particular direction without anyone explicitly commanding them to do so.

A company that earns money from usage time will build differently from one that earns money when a task is completed.

A company that needs as much personal context as possible will handle data differently from one that makes a technical commitment to limitation.

A company whose social responsibility begins only after profits are distributed will make different decisions from one that builds impact into its product and value cycle from the outset.

FlameP cannot claim to be different merely because we write the right sentences on a website.

We have formulated rules that also apply when they become economically inconvenient.

Our manifesto sets out three principles for this: evidence before trust, completion before retention, and value flow before extraction. They are not intended to be mere communication. They are visible in the product, data structures, metrics, and approvals.

4. Evidence Before Trust

Companies often like to talk about trust.

Trust sounds warm. It sounds human. It also spares us uncomfortable details.

We do not want people simply to have to trust FlameP.

They should be able to see which context was used for a task, what was excluded, which provider was involved, which storage rule applied, and whether a connection was closed again after completion.

That sounds technical, but it is actually a social question.

Anyone who asks a person for trust without giving them the ability to exercise control distributes power very unevenly.

FlameP will not conceal this power behind a friendlier interface. We want to make the relevant processes traceable. Not every technical detail has to be shown to every user at all times. But the essential decisions must not remain invisible.

People should be able to understand what happened to their task and their context.

Our users should not have to become data-protection experts themselves, but trust without the possibility of verification very quickly becomes just another word for dependency.

Our personal manifesto therefore contains a sentence against which we must be measured:

Key ideaYou do not simply have to trust us. You can see what happened.

This is not yet a fully fulfilled product promise. FlameP is under development. But it is the direction guiding the technical architecture we are developing.

5. Completion Before Retention

A person’s time is not a raw material either.

Many digital products are assessed by how long people stay, how often they return, and how many interactions they generate. This produces endless feeds, autoplay, artificial streaks, and notifications whose only purpose is to trigger the next use.

The European Commission now explicitly regards such mechanisms as a potential source of systemic risks. In preliminary findings concerning TikTok, as well as Instagram and Facebook, it cites infinite scrolling, autoplay, push notifications, and highly personalized recommendation systems as elements of potentially addictive design. The proceedings have not yet been finally concluded. But they show that product design has long ceased to be treated merely as a question of convenience or taste. [3] [4]

FlameP must not benefit economically from a person never feeling that their concern has been sufficiently resolved.

When a task is complete, the system must be able to stop.

The result is saved or deliberately discarded. Open issues remain visible. Temporary permissions expire. Connections are closed. After that, the person may leave.

No automatically generated next task.

No endless stream of further possibilities.

No notification merely to improve a metric for daily returns.

That does not mean people should use FlameP only once. Life constantly generates new tasks, texts, decisions, and projects. We must be helpful enough that people choose to return.

But the old task must be allowed to end.

FlameP does not want people to stay longer. FlameP is intended to help them move forward more effectively. The Product Integrity Contract already translates this aspiration into planned product objects and metrics, including task completion, time to clarity, closed routes, and deliberately chosen resumption.

In some places, it would probably be economically easier to prolong use artificially.

We know that.

But not everything that improves a metric also improves the company we want to build.

6. Value Flow Before Extraction

FlameP’s social dimension does not begin with a later donation.

It begins with the question of how value is created and distributed in the first place.

If FlameP gives a person time back, that is value.

If their context remains under their control, that is value.

If someone gains access to information in their own language, that is value.

If good work is paid fairly, that is value.

If clear processes enable a company to make better decisions, that is economic value.

And if FlameP generates revenue from this, financial value is created.

Part of it is needed for operations, development, security, and reserves. Without this foundation, any long-term social commitment would be unserious.

But the value created must not end up entirely with FlameP, the connected models, infrastructure companies, and perhaps eventually investors.

UN Trade and Development warns that artificial intelligence infrastructure, capabilities, and economic benefits may become heavily concentrated among a small number of companies and countries. The organization therefore calls for an approach in which access, data, infrastructure, and capabilities are distributed more broadly. [5]

For FlameP, this does not lead to a blanket promise to donate an arbitrary percentage.

Such a sentence would be quick to write. At present, it would not be defensible.

We want to build a traceable value cycle. It must distinguish whether an amount has merely been calculated, reserved internally, committed as a binding obligation, actually paid out, or confirmed by evidence. Reserved does not mean paid out. Committed is not yet confirmed impact.

Admittedly, that sounds somewhat like accounting rather than a grand vision. Good.

Social responsibility does not become more credible when it is expressed more emotionally. It becomes more credible when someone can verify what actually happened.

7. The Social Dimension Runs Through the Entire Company

Guatemala is a concrete test case for us. But it is not the entirety of FlameP’s social dimension.

The social dimension is already present in how we treat users.

Whether we respect their context as their property.

Whether we give them back their time.

Whether we make decisions traceable.

Whether we also take into account languages and lived realities that are commercially uninteresting to large providers.

Whether people participate in applications or serve only as data sources and test groups.

Whether we claim impact only once it is supported by evidence.

Whether we choose not to build a feature even though it promises growth because it unnecessarily retains or disempowers people.

In Guatemala, this becomes more concrete. There, we see how strongly language, education, income, and access determine whether technology empowers or excludes people.

An application in a local language does not solve poverty. It does not replace medical care or automatically create economic independence.

But it can reduce dependency in one specific area.

Admittedly, that is not saving the world.

Nor is it nothing.

What matters is that local people do not become mere scenery in a European impact story. They must have a say in what is needed, what is understandable, and where the limits lie. That is precisely what our personal foundation establishes as a test.

8. We Are Not Saying We Have Already Succeeded

FlameP is under development.

Some parts exist. Others are conceptually defined. Some still have to be implemented, tested, and assessed in real-world use.

That also applies to our social and economic ambitions.

A manifesto is not evidence.

A contract is not a functioning product.

A planned technical control is not yet a demonstrated control.

Planned impact is not yet impact achieved.

That is precisely why we distinguish between principles, concept, implementation, and demonstrable functionality. The Product Integrity Contract establishes the principle that FlameP may promise only what the system actually supports.

That sometimes makes the communication less spectacular.

So be it.

Credibility does not arise from speaking particularly grandly about the future. It arises when we state visibly what already works and what does not.

9. How We Want FlameP to Be Measured

Of course, we will measure user numbers and revenue. Anything else would be entrepreneurially blind.

But these figures must not determine on their own whether FlameP is successful.

We also want to know whether people were able to complete their tasks meaningfully. Whether they reached a clear decision more quickly. Whether less context was shared than would have been technically possible. Whether a connection was terminated cleanly. Whether time was given back. Whether social value actually reached its destination or appeared only in a presentation.

We want economic success and social responsibility to keep each other in check.

The business must be strong enough that protection, development, and social programs do not depend on chance.

Social responsibility must be strong enough to prevent the business from eventually serving only itself.

We will probably make mistakes along the way. No, we will certainly make mistakes.

What matters is whether we recognize them as deviations or eventually declare them the new normal.

10. Conclusion: Our Commitment as Founders

We do not want to build a company that places itself morally above others.

FlameP will face the same economic temptations as other technology companies. More data can improve results. More use can increase revenue. Bigger promises can attract attention. Later corrections are easier than saying no today.

That is why good intentions are not enough.

We need rules, evidence, and limits.

And we must be prepared to defend them even when they become inconvenient.

We want to build FlameP so that people do not have to believe in us, but can examine our decisions.

We do not want to pit economic success against social responsibility.

We do not want to conceal missing functionality with grand words.

And we want to retain the courage not to build a feature when its real value lies in holding people for longer, collecting more information about them, or taking away a decision that should remain theirs.

We simply do not want to build a company that turns people into raw material.

We want to build one that takes work off people’s hands without taking away their time, context, judgment, and share of the value created in return.

That is how FlameP should be measured.

Not by this article.

By the company that emerges from it.