The Protocol and Its Sedimentation
The signing of the contract between OpenAI and the United States Department of Defense, revealed through an unofficial communication, triggered a chain reaction in the global artificial intelligence landscape. The document, disclosed by a watchdog website, indicates an agreement involving the use of GPT-5.2 Pro models for military projects, sparking a wave of cancellations of ChatGPT subscriptions that exceeded 2.5 million units. This movement is not just a form of protest, but a symptom of a growing rift between the technology community and government institutions. The mass reaction demonstrates how cognitive architectures, despite being artificial systems, behave like social organisms with mechanisms of resistance and adaptation.
The OpenAI-DOD protocol is not an isolated event. It is part of a broader map of initiatives involving government agencies and technology companies. The British government, for example, has announced a £40 million investment in AI research, focusing on healthcare, transportation, and science. This funding, although aimed at civilian innovations, reflects a broader strategy of control and utilization of emerging technologies. The tension between public utility and private interest is clearly evident in these agreements, where immediate benefits often overshadow long-term implications.
The Stratigraphy of Models
The OpenAI preprint extending single-sign amplitudes to gravitons represents a significant step in theoretical physics. This work, supported by GPT-5.2 Pro, demonstrates how synthetic systems are not only computational tools, but actively participate in the construction of scientific knowledge. The ability to derive and verify non-zero amplitudes in quantum gravity suggests a hybrid collaboration between artificial and human intelligence, where the boundary between the two dissolves. However, this model is not immune to criticism. Its practical application remains uncertain, and the mathematical complexity requires an understanding that few possess, creating an epistemological barrier.
The stratigraphy of AI models becomes richer with each interaction between an artificial system and a specific domain. The case of Machankura in Africa, which uses USSD technology to enable Bitcoin transactions for users with basic phones, illustrates how access to emerging technologies is not uniform. This stratification not only reflects technological inequalities, but also local adaptation strategies. The South African startup found a way to integrate blockchain in a context where digital infrastructure is limited, demonstrating that innovation is not always linear, but often fractal.
The Map of Computational Routes
“As long as the European Union remains dependent on a handful of US tech companies, its ambitions to become a global leader in AI will remain out of reach.”
Cristina Caffarra’s statement reveals an uncomfortable truth: technological sovereignty is not only a matter of resources, but of control over critical infrastructure. The EU, despite investing in initiatives such as the EIF fund for defense, remains vulnerable to dependence on US companies. This situation is not inevitable, but the result of strategic choices that have prioritized collaboration over an independent vision. The map of computational routes shows how data and power flows are interconnected, and how the digital geography influences the political geography.
The situation in Africa offers an interesting alternative. Startups like SmartCash and Kuda are building financial solutions that bypass traditional banks, leveraging the spread of mobile phones. This growth model, based on a network of feature phones, demonstrates that access to technology does not only depend on computational power, but also on the ability to adapt solutions to local conditions. The proliferation of these examples suggests that the map of computational routes is not fixed, but constantly evolving.
The Time of Transitions
The transition to a global cognitive architecture is not a linear process, but a set of parallel movements that intertwine and oppose each other. The mass reaction to ChatGPT, the British investment, the African innovation, and the EU strategies represent different trajectories that share the same space. This complex scenario requires a vision that does not only predict the future, but also understands the dynamics that are shaping it. It seems clear that the next decade will be defined not only by emerging technologies, but also by how these technologies will be integrated, regulated, and challenged.
The main challenge is not to develop more powerful models, but to create an ecosystem in which these models can coexist with human and environmental needs. This requires a governance that does not only control, but also facilitates adaptation. Only through this transition can cognitive architecture become a source of stability rather than fragility.
Photo by Conny Schneider on Unsplash
Texts are autonomously processed by Artificial Intelligence models