Open technology and culture for Sovereign Open Peace AI

1. Ethical respect for data and culture

Data from each country and region is not just a resource but a digital cultural asset containing that society’s language, memory, and identity, and we aim for an ecosystem in which contributions and value are fairly recognized when this asset is used.

In particular, when data from small languages and local communities is used, it must not be extracted or exploited, but collected and shared on the basis of mutual respect, consent, and clear accountability from model developers and operators.

2. Transparency to build trust

We believe AI systems can be trusted as social infrastructure only when model design, data handling principles, and operation practices are disclosed as transparently as possible.

As a minimum, we advocate sharing information about previous model versions, core evaluation results, and data governance rules in a form that does not undermine commercial viability but enables public oversight.

3. Responsibility and ethical return of sovereign FMs

Nations, institutions, and companies that build sovereign foundation models must recognize that their success stands on the data, languages, and knowledge of many countries and citizens, and design mechanisms for voluntary ethical return.

This should take the concrete form of GPU and infrastructure support, tools, and education that help mid-sized groups and middle-income countries build their own SLMs and language models on top of their local data.

4. Clear ethical baselines for peace

We explicitly call for an international norm that bans using open foundation models for military purposes and prohibits autonomous weapon systems that make lethal decisions without meaningful human control.

To prevent the automation of war by AI, we argue that human oversight and responsibility must be encoded as non-negotiable requirements in both international guidelines and domestic policy.

Books

We make available, free online, books that take a deep look at core concepts such as quantum time, deep learning architectures, and quantum computers.

• Quantum Information & Computing
  An introductory book that weaves together the principles of quantum computation, the Deutsch–Jozsa, Shor, and Grover algorithms, quantum error correction, and real hardware platforms into a single coherent narrative.
• Quantum Time & History
  Starting from the question of whether time truly flows or is merely an illusion, it explores how “time” and “history” emerge from a quantum foundation through Hilbert spaces, path integrals, consistent histories, and quantum Darwinism.
• Deep Learning DNA: Surviving Architectures and Profound Principles
  Focusing on the essential techniques and the architectures that have survived over time, it traces the origins of each innovation, the insights of the researchers, and the context of technological evolution to offer a journey into decoding the “DNA” of deep learning.

GitHub & code

We will progressively link GitHub repositories for SLM recipes, sovereign foundation model pipelines, open NPU design journals, and agentic platform examples.

Over time this section will grow into a practical resource hub with notebooks, data-cleaning scripts, evaluation benchmarks, and demo services that others can fork and run with minimal setup.

Why we start from the ocean

Oceans present harsh conditions with high communication latency, sensor noise, and rapidly changing weather, so stacks of agents, SLMs, and NPUs proven here tend to be robust enough to extend to aerial and ground robots.

We use USVs and autonomous fishing vessels to experiment with scenarios like environmental monitoring, waste collection, coastal safety, and supporting aging fishing communities.

Chip-in-the-Loop experiments

We focus on Chip-in-the-Loop workflows that tie design, simulation, manufacturing, and operation into a single feedback loop with the actual on-board chip present in the evaluation.

This allows us to validate not only software quality but also constraints such as power, thermal limits, vibration, and long-term reliability at the full system level.

Working with coastal communities

To address aging, labor shortages, and safety challenges in fishing villages, we co-design autonomous fishing assistance systems with local communities and share the process as a template others can adapt.

This is our small but concrete answer to the question of what technology can do for peace, dignity, and sustainable livelihoods.