The Golden Zeta: Prime Numbers from the Norm Form of ℤ[φ]

We present two machines for finding prime numbers. The first, due to Riemann (1859), guesses the prime count with the logarithmic integral and corrects with waves from zeta zeros — an approximation that never terminates and whose convergence requires the Riemann Hypothesis. The second uses the golden norm form N(a + bφ) = a² + ab − b², which takes two integers as input and produces an integer as output. Every prime p ≡ ±1 (mod 5) is directly manufactured by this form as a product of golden conjugates: p = (a + bφ)(a + b − bφ). Primes p ≡ ±2 (mod 5) are invisible to the norm form — the golden ring's dark matter — identified by the absence of solutions and confirmed by a simple congruence test. The complete set of primes is recovered exactly: split primes from the factory, inert primes from the Legendre patch. The Riemann zeta function is then expressed as ζ(s) = ζ_ℚ(√5)(s) / L(s, χ₁), where the numerator is the Dedekind zeta function of the golden ring — counting ideal norms from the same form — and the denominator is the Legendre L-function that accounts for the dark matter. In the golden variable w = φ^(−s), the critical line becomes the circle |w| = 1/√φ, and the functional equation becomes golden inversion w → 1/(φw), whose unique fixed circle is |w| = 1/√φ. Since every zero of ζ is a zero of the golden ring's counting function, and the norm form's axis of symmetry (φ ↔ ψ) corresponds to σ = Re(φ) = ½, the zeros inherit the golden symmetry. The primes are not mysterious objects whose distribution must be guessed and corrected. They are the output of a two-integer machine running on φ² = φ + 1.