In the vast galaxy of artificial intelligence, a quiet force has emerged from an alpine outpost. ETH Zurich, a Swiss bastion of scientific precision, in conjunction with EPFL (the vibrant tech institute), and the Swiss National Supercomputing Centre (CSCS).has released a new large language model, and it’s not just another tool in the tech arms race; it is a spark of rebellion in a world increasingly dominated by imperial AI corporate superpowers.

For years, American AI giants have ruled the frontier. With glossy promises of safety, alignment, and progress, companies like OpenAI and Anthropic have built towering models in the shadows, trained on secret data, optimised for engagement, and fine-tuned to please shareholders first and society second. Their creations are immense and powerful, but fundamentally closed, centralised, and appear to extract more than they give.

They are the Empire. Their models are the Death Star — vast, cold, and controlled by a band of ultra-powerful Tech feudalists – and engineered to bend reality toward their narrow definition of safety and success. The most dangerous thing is they seem to think they are the Rebel Allianace.

The Force is growing

But wait – in a moral galaxy far far way (Europe), a pioneering model has been developed. It arrives not with a sales pitch, but with purpose; it is open, transparent, rigorously documented, and rooted in public trust. It is the X-wing slicing a swathe through the Empire’s fortress defence: not built for scale alone, but for understanding, collaboration, and freedom.

While Silicon Valley builds in secret, this LLM builds in daylight. While American models speak in polished PR-approved outputs, this model speaks the language of science — honest, explainable, multilingual, and unafraid of complexity. It does not presume to be the arbiter of truth, it invites scrutiny. It does not dominate, it cooperates.

This is not just a technical divergence; it is a philosophical one. Where the US models are trained to serve platforms and profits, the Swiss model is trained to serve knowledge, inquiry, and public good. It is built not as a product, but as a public utility, not as surveillance infrastructure, but as an educational companion, not as a black box oracle, but as a transparent tool.

The AI seep

And in this moment, that matters. AI is seeping into law, education, media, healthcare, every institution that shapes human life. The models we embed into these systems carry assumptions, values, and incentives. If we are not careful, we risk replacing old power structures with even more invisible ones, coded into the foundations of daily life.

This model resists that. It is not just another LLM. It is a countermodel, a resistance to the normalisation of closed systems and algorithmic opacity. It shows that Europe, and more broadly the world outside the tech monopolies, can offer something better — not just different, but better. Principled, accessible, and aligned not with markets, but with people.

The fight ahead is not one of scale, but of vision. It is not about who builds the biggest model, but about who builds the most trustworthy one. And while the Empire may have more compute, more capital, and more control, the rebels have something it never will.

This model is not the endgame, it is a new hope; one where intelligence is open, power is shared, and the Force is, at last, with the people.