The Paradox of Control: n8n, Automation, and the Hidden Dependency on Big Tech

For decades, the promise of automation was almost religious. Let machines do the heavy lifting, free up human time, allow systems to understand each other without asking for permission. The utopia of total control: a digital orchestra where every instrument knew its part and the conductor could disappear.

But reality, as usual, slipped between the gears.

n8n: an attempt to return automation to the user

n8n (short for “nodemation”) was created in 2019 by German developer Jan Oberhauser, who wanted to design a visual automation tool that didn’t belong to any company. Inspired by platforms like Zapier or Integromat, he built it around nodes, small functional blocks that represent actions such as sending an email, querying an API, or generating a file.

The real innovation wasn’t just the interface, but the idea that anyone could host it on their own server, extend it, and make it part of their own environment. In a world dominated by SaaS subscriptions, that sounded revolutionary.

Over time, n8n grew thanks to its community and its fair-code model, a license that makes the source code public but limits direct commercial use without permission. It was a compromise between the ideal of open software and the practical need to survive in an ecosystem ruled by corporations. But the question remained: can code be truly free if the infrastructure it runs on isn’t?

The dream of free automation

n8n became a symbol of independence in an era obsessed with the cloud. It promised to give users control again, to build their own workflows and execute them without paying rent to anyone.

It was a powerful idea. Yet the paradox appeared quickly. Even when you host n8n yourself, your workflows still depend on external APIs such as Gmail, Slack, AWS, or Google Drive. Each one is an invisible thread tying your system to the infrastructure of the same technological giants you thought you had escaped.

The illusion of independence

The code is free, yes, but execution, traffic, storage, and connectivity all live inside ecosystems controlled by a few companies. It’s like owning your own ship but always sailing through private waters.

Most open-source projects face the same dilemma: technical independence doesn’t mean structural sovereignty. You can build your own system, but if it runs on AWS, your freedom is temporary.

This isn’t a flaw in n8n; it’s a symptom of the modern ecosystem. We live inside a network of dependencies so efficient that it disappears from sight. Netflix understood it before anyone else. When it outsourced its entire infrastructure to Amazon Web Services, it helped shape the cloud model that now defines the world. What looked like agility was really a transfer of power. Infrastructure became a service, and control turned into rent.

Comfort as contract

Every act of automation carries a pact. To automate is to give up intervention. In that gesture lies the price of control.

When a developer designs a workflow in n8n, they rarely think about the chain of systems that sustain each action. A “send email” step seems harmless until you remember that it passes through Google’s servers, Amazon’s networks, and layers of protocols few people can even audit.

Free software gave us the illusion of choice, but the global infrastructure had already decided where the data must flow. Control stopped being an action and became a configuration.

When the machine starts to decide for you

n8n allows you to automate almost anything. It can read and label emails, send automatic replies, create tickets, schedule reports, and trigger alerts without you moving a finger. That power, which looks like freedom, hides a quiet surrender.

Each time you let a workflow decide what matters, what to answer, or what to store, you give up a fragment of your own judgment. We automate because decision-making is exhausting, and in that exhaustion we stop noticing. Some tasks, like reading your own messages or replying yourself, are worth keeping precisely because they are human. They take time, yes, but they preserve nuance.

The danger of automation isn’t that machines will think for us; it’s that we’ll stop thinking because they already do.

Towards sovereign automation

The question isn’t whether we should stop automating, but how to reclaim sovereignty over the process. A new generation of architectures is beginning to explore that path: local automation, edge computing, lightweight language models that process data on the device instead of the cloud, and federated systems that share workflows without a central authority.

A truly autonomous n8n isn’t a fantasy. Imagine a network where workflows learn locally, instances communicate directly, and systems grow without a single point of control. It wouldn’t just be a technical milestone, but a political one. Real autonomy doesn’t depend on who writes the code, but on who owns the ground it runs on.

Epilogue: control as a contemporary illusion

Real control doesn’t live in the software, but in the infrastructure beneath it. As long as that infrastructure belongs to someone else, every act of automation will remain a disguised surrender.

The paradox of our time is not our dependence on technology, but how natural that dependence feels. We believe we are free because the algorithms obey, but every command we give strengthens the system that governs us. We automate to save time, to be more efficient, but in doing so we train the machines to think in our place.

Perhaps the future doesn’t depend on creating more automation, but on learning to look beneath it, to see which part of our control is real and which was, from the beginning, a beautifully programmed illusion.

R.Adams

Glossary

n8n: an open-source visual automation tool created by Jan Oberhauser that allows users to connect services and APIs through workflows made of nodes. It can be hosted locally or in the cloud.
Fair-code: a licensing model between open source and proprietary software, where the code is visible and modifiable but cannot be used commercially without permission.
Edge computing: a distributed model that processes data near the source or device rather than in centralized servers.
LLM (Large Language Model): an AI system trained on large text datasets, capable of generating and understanding language, sometimes integrated into automation workflows.
Federated systems: decentralized networks where independent instances collaborate and share information without relying on a single central authority.