🎈ry

The French government was not in exile at all. Never. The French government was never in exile. It was very much there.

The Greens are a parliamentary group, not a sociological community. What they vote on has a concrete political impact.

Bye.

"Europeans Greens"

Take off your little antifascist badge if you support EU.

If you think naming potential use equals liberation, we’re not in the same conversation.

Bye.

You’re not wrong that people will always find cracks to speak through. But that is not a feature of the tool. That is a testament to human resilience.

This technology is trained on unconsented data, built on extractive labor, embedded in centralized systems, and shaped by corporate logic. Using it “against” the system often just feeds it new data, new forms, new edge cases to neutralize.

You’re confusing potential with structuree. The fact that a tool can be used to produce subversive knowledge does not mean it was designed for that, or that it supports it.

You can’t separate a tool from the infrastructure, labor, extraction, and epistemicide that make it possible.

Usefulness is not neutral. It reflects what the system rewards, what it makes legible, and what it demands from those forced to live under it.

Wanting to tear it all down while clinging to the usefulness of LLMs is like calling for abolition while defending predictive policing because it helps plan public transport.

So, if, your defense is that people resist anyway, you’re describing survival. Not freedom.

And yeah, people have always used the tools of colonialism to resist. That doesn’t make the tools neutral. It proves how resilient oppressed communities are in hostile systems. Resistance through the master's tools is often necessary, but never liberating on its own.

You’re still missing the point. Saying LLMs flatten thought is not the same as saying marginalized people lose all capacity to think. It means the tools they are increasingly pushed to use are structured to filter, erase, and reshape their language in ways that align with dominant norms.

What matters is not what is technically possible. What matters is what is materially accessible and under what conditions. That is the real terrain of power.

The fact that a privileged minority can modify these systems does not make them open. It makes them selectively porous in ways that still serve capital accumulation and intellectual enclosure.

Oui, some people can. But that is not general access. Training or fine-tuning an LLM requires compute, infrastructure, and expertise that remain concentrated in elite institutions, corporations, and a narrow technical class.

You do not resist power simply by using the master's tools.You resist by confronting the conditions that made those tools necessary, by refusing the framing, and by building outside it.

Saying knowledge and power are inseparable is not an excuse to ignore how much of today’s “resistance” is preformattted, captured, and recycled through tools built to neutralize it.

Yes, dictionaries reflect dominant ideology. The point is not that they cannot be used. It is that their form encodes hierarchies, exclusions, and norms born of specific histories of power.

Claiming it's all about how we use it erases the material conditions and structural violence embedded in the tool itself. That is not nuance. That is liberalism pretending to be realism.

Yes, people can resist and repurpose. But resistance does not prove neutrality. It proves that people are creative even under constraint. The model remains designed to serve corporate interests, to filter language, to prioritize what isprofitable and palatable to power.

Saying it depends how you use it assumes technology is neutral. It is not. LLMs are built, trained, and deployed within structures shaped by capital, surveillance, and ideological control.

It's the grammar that gives those other drives form and coherence within the current system. So when people say "it's about profit," they may not be saying everything, but they're pointing in the right direction.

You're right that other drives are at play: control, prestige, even delusion. But profit isn't some surface-level motivation we can dismiss.

That doesn't mean it's only about short-term profit. Often it's about speculation, dominance, ideological influence, and long-game positioning. But these are all subsumed under the horizon of capital accumulation.

It's about how every move, every model deployment, every strategic blunder is still shaped by the imperative to capture markets, preempt regulation, and consolidate CONTROL.

I get where you're coming from, but I think you're underestimating how profit operates not just as a motive but as a structural logic. It's not about individual actors making perfect business decisions.

So YES, the prioritization of profit and control is embedded in the design, training, and deployment of these models from the ground up. It is impossible to separate the algorithm from the capitalist structures that fund and direct it.

The whole architecture is a form of digital gatekeeping that reproduces existing power hierarchies under the guise of “safe” and “helpful” AI.

This means that the algorithms are not neutral linguistic tools but instruments of surveillance, controll, and capitalist accumulation. The profit motive shapes what knowledge, which perspectives, and whose voices get amplified or erased.