Assuming systems behave similarly at all sizes is flawed. • Two assumption types to avoid: load and interaction. • Larger scales involve different physical stresses and social dynamics. • Verify scaling through testing, not assumptions.