zlacker

If you really think that what you're working on poses an existential risk to humanity, continuing to work on it puts you squarely in "supervillian" territory. Making it closed source and talking about "AI safety" doesn't change that.