Personal tech beliefs

I care about data sovereignty: people and communities should know where their data lives, who can see it, and what happens when it moves. Systems that obscure that chain erode trust.

Access to information should be as transparent and reachable as possible for anyone who legitimately wants to understand or use it without trapping people in black boxes or vendor lock-in as the default.

That openness has to sit alongside clear ownership: not everything is "public by default," and consent, context, and purpose still matter.

If data or creative work has value, the people who produced it deserve recognition and fair compensation when others commercialize or build on it extraction without return is a design choice, not a law of nature.

Individuals should stay responsible for what they upload, consume, and interact with: ignorance is not a permanent excuse when tools make choices easy to automate. We can build interfaces that nudge clarity instead of dark patterns.

I treat AI as a utility for acceleration and exploration: summarizing, drafting, checking, and surfacing options always with human judgment, accountability, and domain context in the loop.

Treating models as a full replacement for people, with no supervision or guidance, is a recipe for brittle systems, ethical drift, and silent failure. The goal is leverage, not abdication.

Ethical data consumption means minimizing what we collect, being honest about why we need it, and retiring it when the job is done not hoarding by default.

Infrastructure should be designed with environmental cost in mind: efficient workloads, honest capacity planning, and avoiding performative complexity.

Technology should be implemented in ways that reduce harm: non-violent in the broad sense no exploitation of users, workers, or ecosystems as an acceptable "optimization." That's the bar I want my work to meet.

I will not integrate AI or my labor into systems whose primary purpose is mass surveillance of people: dragnet collection, pervasive facial or behavioral scoring, or other infrastructure that treats whole populations as a default suspect class. Scale and "efficiency" are not excuses to normalize that posture.

I also do not participate in work that embeds AI into offensive military integration: systems built to automate killing, degrade human judgment in the use of force, or optimize violence. There is a world of difference between defensive safety and tooling that makes harm faster, cheaper, or less accountable.

AI and infrastructure should be aimed at improving life health, education, accessibility, creative practice, climate-aware operations, and resilient communities not at taking it or at stripping dignity from people at scale. If a project drifts the other way, the right answer is to stop, speak up, or walk away.