On Information Factories – Article published on Akureyri.net

I'm not launching a crusade to change how people speak. But I do believe the term artificial intelligence is both misleading and emotionally loaded. That’s why I’m trying something different: thinking of this technology not as “intelligence” but as an information factory.

The myth that clouds the conversation

Think about what “artificial intelligence” evokes. Hollywood machines like HAL 9000, Skynet, or friendly humanoids with consciousness and will. These myths, shaped by decades of science fiction and tech industry hype, dominate the conversation. They fuel both fear and misplaced faith—and blur the actual stakes.

The reality? We’re not dealing with sentient beings. We’re dealing with industrial tools. And it matters that we see that clearly.

So, what is an information factory?

Instead of intelligence, picture a factory.

This factory processes text and data instead of raw materials. It doesn’t understand meaning—it computes probabilities. It produces text, code, images—outputs that may look coherent but can contain hallucinations and errors.

We, the users, have become the quality control. If we don’t take that responsibility seriously, we risk being misled by flawed or synthetic content.

Why does this metaphor matter?

Thinking in terms of factories has four key consequences:

1. It breaks the myth. We stop talking about sentient beings and start discussing tools.
2. It clarifies accountability. Who owns the factories? Who controls the inputs and outputs? Who profits?
3. It empowers users. We stop asking what the “AI believes” and instead ask: What does this factory produce when I input this data?
4. It reveals the political scale. A handful of foreign-owned information factories now control much of the global output. If we base schools, institutions, and businesses entirely on their services—without building our own infrastructure—we forfeit autonomy.

Icelandic factory or digital dependency?

This isn’t just an academic point. It’s about digital sovereignty.

We need our own information factories—trained on Icelandic data, aligned with Icelandic values. Not to isolate ourselves, but to preserve choice and agency.

Calling things by their right name

This is an experiment—not a doctrine.

But if we want to have grounded, responsible conversations about this technology, we need to choose our words carefully. Not out of dogma, but out of clarity.

If we stop seeing these machines as “intelligent agents” and start seeing them as information factories, we’ll be in a much better position to use them wisely.