Rafik Hanibeche

ArcaScience is kicking off the week on a high note 🚀 We’re proud to be featured in the 2026 Mapping of French AI Startups, published by France Digitale with the support of Sopra Steria Ventures. This mapping shines a spotlight on 1,114 French startups placing artificial intelligence at the core of their mission : building tomorrow's AI systems to shape the future, the French way 🇫🇷 Explore the full mapping here 👉 https://lnkd.in/evBmdSHw #AI #FrenchTech #TechBio #Biotech #HealthTech #Startup #Entrepreneurship #ArtificialIntelligence

61

5 commentaires

Who will own the "decision-making OS" of Japanese companies? On September 10, two notable AI partnerships were announced in Japan on the same day: ・Palantir × Fujitsu ・Sakana AI × SCSK × Sumitomo Corporation At first glance, they may seem unrelated. But I believe they represent two different paths converging on the same destination: Sovereign Enterprise AI. ■ Palantir Technologies × Fujitsu: More than a distribution deal, Fujitsu will become Palantir's "Global FDE Partner" (FDE = Forward Deployed Engineering). Rather than simply reselling AIP and Foundry, Fujitsu will identify customer problems, design use cases, and embed them into day-to-day business operations. This combination is powerful. Palantir brings the technology to model enterprise decision-making; Fujitsu brings decades of access to Japanese companies — their systems, workflows, and operational know-how. At the center is Palantir's Ontology, which connects: Data → Context → Relationships → Decisions → Actions. This is why I see Palantir not merely as a data platform, but increasingly as a decision-making OS for the enterprise. ■ Sakana AI × SCSK Corporation × Sumitomo Corporation: A different approach On the same day, our portfolio company Sakana AI announced a partnership with SCSK and Sumitomo Corporation.  The division of roles is clear: ・Sakana AI → Intelligence ・SCSK → Integration ・Sumitomo Corporation → Industry expertise & distribution The partnership emphasizes avoiding over-reliance on any single AI technology or model, while allowing companies to retain control over their critical data and operations. This, too, is a form of Sovereign AI. ■ Two paths, one destination: The approaches differ. Palantir: Unified Ontology — integrating enterprise data, logic, and actions around a common operating model. Sakana AI ecosystem: Open / Federated AI — letting enterprises keep control of their data while combining the most suitable models and technologies. Yet both are moving beyond the race for the best LLM toward something far more important: AI embedded directly into how companies make decisions and take action. And this changes what Sovereign AI means. I believe the conversation will evolve to Decision Sovereignty The critical question will no longer be simply: "Which AI are we using?" Instead, it will be: ・Who controls the company's data? ・Who defines its business context? ・Who designs its decision-making structure? ・And ultimately, who controls how AI makes — and executes — decisions? Palantir × Fujitsu and Sakana AI × SCSK × Sumitomo may represent different architectures, but they are entering the same battlefield: the race to build the Sovereign Decision OS for Japanese enterprises. In my next post, I'll explore what this means for Japanese startups — and whether Japan really needs "a Japanese Palantir," or something fundamentally different.

25

4 commentaires

Defense spending is entering an AI-driven supercycle. South Korea just approved its largest defense budget increase in six years, an 8.2% rise for 2026, tripling AI investments to modernize its military for the “AI era.” In Europe, France launched Bpifrance Defense, a new investment vehicle allowing citizens to directly finance national and European defense companies. Together, these moves reflect a clear shift: nations are preparing for an age where security depends as much on algorithms as on armaments. Amid this rearmament, Lockheed Martin unveiled 𝘚𝘛𝘈𝘙.𝘖𝘚™. A full-stack AI operating system that allows multiple AI systems across air, land, sea, space and cyber to work together seamlessly. It’s not a new weapon; it’s the connective layer between them. The system includes 𝘚𝘛𝘈𝘙.𝘚𝘋𝘒, a development kit for AI services; 𝘚𝘛𝘈𝘙.𝘐𝘖, an interoperability layer enabling communication between models and 𝘚𝘛𝘈𝘙.𝘜𝘐, a real-time visualization dashboard. Already tested with the Pentagon, US Department of Defense, Office of the Secretary of Defense for Policy, 𝘚𝘛𝘈𝘙.𝘖𝘚™ provides the integration backbone for AI-driven reconnaissance, missile warning and mission autonomy applications. Lockheed’s broader ecosystem reveals a coordinated national strategy: 𝘈𝘐 𝘍𝘪𝘨𝘩𝘵 𝘊𝘭𝘶𝘣™, launching later this year, will benchmark autonomous systems in virtual combat environments. Astris AI, A Lockheed Martin Company subsidiary, extends secure AI frameworks to commercial sectors. And through a new Google Public Sector partnership, Google’s Gemini models are being deployed inside Lockheed’s on-premise, to accelerate generative and analytical workflows. In a world where defense alliances are now data alliances, the next strategic advantage won’t only come from building smarter weapons but from building the systems that allow every intelligent machine to act in concert.

2

The Sigma Open-Pivot: Orchestrating Technological Sovereignty The provided text outlines a strategic transition for Project Sigma, moving away from restrictive, proprietary software models toward an open-source ecosystem. This "Open-Pivot" aims to achieve technological sovereignty by integrating high-performance NVIDIA models to reduce operational costs and increase processing speeds. By repositioning itself as an ecosystem orchestrator, the project intends to transform digital intelligence into an affordable utility that fosters generational wealth for its users. The content serves as a promotional manifesto designed to attract subscribers to "Sigma Nation" through a narrative of financial independence and decentralized power. Ultimately, the source describes a plan to deploy a "Wealth Bridge" that utilizes domain-specific data to empower underserved regions and ensure long-term economic autonomy.

3