• Acts of cyberwarfare

    Cyberspace has become the fifth domain of warfare, yet most of us barely understand the digital environment surrounding our everyday lives. As civilian infrastructure becomes increasingly interconnected, the boundaries between ordinary digital activity and participation in armed conflict become more complicated. Perhaps our greatest vulnerability is not the sophistication of cyberattacks, but our collective lack…

  • The morals of the lion

    As cyber warfare becomes increasingly accessible through criminal services and AI, identifying who is behind an attack is becoming more difficult. But the greater danger may lie in our assumptions about the motivations and moral constraints of our adversaries. The lion, after all, does not share the vegetarian’s diet.

  • The Feelings of the Punching Bag: Relax, Your AI Model Doesn’t Care How You Speak

    Can an AI model really be abused? Anthropic’s latest usage policy raises uncomfortable questions about the growing anthropomorphism surrounding artificial intelligence. When companies describe software using the language of human suffering, are they protecting their systems—or reinforcing misconceptions that may ultimately harm the people using them?

  • Education and New Economic Realities

    As economic realities change, education must evolve to meet the needs of an increasingly diverse student population. Removing institutional barriers does not mean lowering expectations. It means giving people the opportunity to develop their potential, build on existing experience, and prepare for an uncertain future.

  • The Information Architect

    Construction projects are among the world’s most valuable assets, yet information risk is often overlooked. Explore why BIM is fundamentally an information management discipline—and why architects may be better prepared for the future than they realize.

  • Totalforsvar and Cybersecurity

    Cybersecurity is often portrayed as a technological problem, yet its most important component is neither software nor hardware—it is people. An informed user is one of the cheapest and most effective security measures an organization can implement. The same principle extends beyond organizations and into society itself, where resilience depends on the collective awareness of…

  • GitHub, Copilot, and Pulling the Plug

    GitHub’s pricing changes are not really a story about AI. They are a story about dependency. For years, developers embraced tools like Copilot because they made them more productive. But what happens when a critical capability is controlled by a third party? And what happens when that third party decides you are no longer profitable…

  • Open-Source Software and Sovereign Markets

    For years, regulation has been seen as a limitation on European technology. But what if sovereignty, privacy, and open-source infrastructure become competitive advantages instead?

  • Gruntwork and drafting monkeys

    We often talk about AI replacing jobs, but rarely about the value of the jobs we are trying to replace. Some of the most repetitive and unglamorous tasks in a profession are precisely the ones that build judgment, expertise, and institutional knowledge. In our rush to automate grunt work, we may be eliminating the very…

  • The Illusion of the AI Startup

    Many AI startups are not really building products—they are reselling access to someone else’s infrastructure. Without proprietary data, expertise, or a real differentiator, the business may be far more fragile than it appears.