Machine-speed decision-making outpaces human deliberation by a lot. Corporate AI operates in a faster time-register than democratic governance can even occupy; which is a problem.

If AI moves faster than societies can respond, governance isn’t going to fix this. We’re talking about a sovereignty problem. Ownership. The big questions.

Novels. Paintings. Research papers. That social media post you made at 2 AM that you kind of regret. All of it … centuries of people thinking hard about stuff, both serious and strange … and now all of that cognitive work is just sitting there like raw material waiting to be extracted and repurposed. These cognitive artifacts are the residue of thought. Nobody’s preserving them in a museum. They’re in massive data warehouses, waiting to get processed.

When AI consolidates all that cognitive work, it has to choose. It picks what’s predictable and efficient, and it sidelines everything else. Homogeneity scales the easiest.

Just to be clear, AI isn’t a neutral tool. It’s not like a really smart hammer. It’s corporate infrastructure: geological extraction, water consumption, human labor, centuries of collected knowledge. That’s what it’s built from.

And these systems don’t think. They automate the data warehouse. They redistribute cognitive labor at speeds no democratic institution can match; not because they’re brilliant, but because the cognitive work that humans already finished has been industrialized.

Who benefits from the cognitive “work” AI performs?

Who pays the material costs?

Who pockets the profits?

 

 

Research Nexus: Political Economy of AI · Dead Cognitive Labor · AI Infrastructure · Critical AI Studies · Data Temporalities