AI, DeepSeek, and hacker ethics: A new era for technopolitics?

The emergence of artificial intelligence tools such as DeepSeek prompts us to reflect on how technological development models are being designed and what their impact or contribution to society might be. These models invite us to revisit hacker ethics, or the values that defined technological development in the late 1950s at the Massachusetts Institute of Technology (MIT).

Closer to the present, in the 2000s, hacklabs emerge—spaces for organization and collaborative learning, often intertwined with social centers and cultural hubs. These hacklabs become a reality where the community, free exchange, and experimentation drive real dynamics of coexistence with the neighborhoods that host them. It is here that people with interests in science, new technologies, digital or electronic arts, culture, etc., converge, sparking social innovation processes and the later emergence of what came to be called citizen labs.

It was in these spaces (hacklabs and social centres) that a significant part of the social mobilization behind the 2011 square occupations took root, mobilizations whose influence lasted until nearly 2020.

Today, when authoritarian movements leverage the breakneck speed of communication as a weapon to clash with our reality and overwhelm us, revisiting discussions about how technology can either widen inequalities or help democratize access to knowledge is crucial. We must ask whether it can foster free exchange and enable the collective creation of more efficient, affordable solutions. If technology can indeed be a tool for social justice, the principles of hacker ethics become our guiding light.

Decentralization and free access to knowledge

One of the pillars of hacker ethics is the belief that knowledge should be accessible to everyone. Decentralization suggests that information, resources, and decision-making should not reside solely with a select few but should be distributed as widely as possible.

Open models, developed by hacker communities and the free software movement, have shown that decentralization can be key to creating technological development that is not only fairer but also more efficient and autonomous.

DeepSeek, with its focus on open-source models, exemplifies this by enabling continuous improvements that aim to democratize AI use. The “DeepSeek revolution” is not just a geopolitical matter (although it has been portrayed that way), but rather an issue of monopolies, offering an alternative to the closed, pay-per-use models of certain megacorporations. (Keep in mind that OpenAI is backed by Microsoft, which for decades has maintained both citizens and governments captive to its operating system, a product often deemed inferior and more costly than Linux.) Instead, DeepSeek adopts a community-driven model that frees not only the AI itself but also the methodology behind it (the truly groundbreaking part). Thanks to this, we have witnessed a qualitative leap in low-cost AI training, with results comparable to those of proprietary models.

By centring on a pure Reinforcement Learning approach, DeepSeek opens new research horizons in which models generate “untutored” Chains of Thought (CoT) and unexpected solutions, precisely what should be at the core of AI development. It also allows for lower-cost training with a smaller carbon footprint, critical in our era of climate emergency and intensifying struggles over energy and computing resources. This shift is vital to avoiding corporate or imperialist monopolies. Moreover, open-source development and reduced training requirements continue to spark new, promising approaches.

Experimentation and collaboration: engines of social progress

Within hacker ethics, experimentation is a mechanism for continuous improvement. In this spirit, AI becomes an arena of perpetual iteration and collective learning. By enabling communities to modify and adapt open models (such as DeepSeek) to their specific needs, the potential of AI expands well beyond commercial applications. This is not only due to broader access: by sharing data and decision-making, a richer diversity of perspectives is incorporated, lessening the biases that often emerge in centralized systems. Furthermore, distributing the computational load can help optimize energy use and shrink the carbon footprint of large data centres.

This method also promotes interdisciplinary collaboration: when programmers, sociologists, educators, and activists pool their knowledge and efforts, innovative solutions arise for complex social challenges, ranging from improving educational processes to enhancing accessibility for vulnerable groups.

The power of technology to transform society

Hacker ethics maintains that technology is not neutral; it can be used to perpetuate entrenched power structures or to challenge them. Decentralization, collaboration, and shared knowledge are essential for the latter. Consequently, the development of artificial intelligence must be guided by these same principles. To counterbalance the power of major corporations, it is important to consider the role of the State in developing public digital infrastructures and, specifically, open and transparent public AIs aimed at solving problems, upgrading public services, and ensuring that advanced tools are available to public administrations, universities, citizens, and businesses alike.

ALIA stands out as a pioneering initiative in the European Union, designed to provide a public AI infrastructure that includes open, transparent language models incorporating co-official languages. It is the first European public, open, and multilingual infrastructure, trained at the Barcelona Supercomputing Center under the auspices of the Spanish government. While there has been notable scepticism about its quality (partly attributable to general disillusionment with public projects and partly to a certain “know-it-all” tone on social media), we should remember that this is a first version, presented as a public model, with particular emphasis on linguistic inclusivity. Though it can (and will) be improved, it positions Spain and Europe, once again, at the forefront of technology in the service of citizens, reinforcing the State’s independence and expanding its technological capabilities.

The advent of open models like DeepSeek or ALIA helps level the playing field when it comes to accessing advanced tools, contributing to a fairer society in which knowledge and innovation are no longer monopolized by a privileged few. Achieving social justice in technology is never automatic, it must be a collective effort.

By aligning with the pillars of hacker ethics, decentralization, experimentation, collaboration, and free access to knowledge, we can begin to imagine futures where AI is not just powerful but also equitable and accessible to everyone.

Special thanks to those who helped refine this article: Gala Pin, Ana Méndez de Andés, Guillermo Zapata, and Carlos Tomás Moro.