Silica is a functional systems programming language that answers the question, “what languages should look like when modern hardware, actor-based concurrency, explicit effects, memory safety, and tool-readable intent are treated as first-class design problems.” This presentation introduces Silica’s premise: many modern languages still carry assumptions from older machine models, leaving concurrency, distribution, side effects, and FFI boundaries to convention rather than the compiler. We will look at Silica’s integrated model of effects, actors, and regions; its stack-shaped, region-owned actor memory direction; its deliberately strict compiler rules; and its “dangerous by design” FFI boundary. The talk closes with the roadmap toward Silica, Project BEES, and Organism as runtime-facing systems built on the same substrate.