[math-fun] Twin Prime conjecture

Zhang's recent paper asserts that there is some integer below 70 million such that infinitely many pairs of primes differ by that integer. Does anyone know if this theorem is generally accepted, (or can we assume that its publication in the Annals is sufficient evidence that it's correct) ? And, was it previously unknown that there exists any integer N such that infinitely many pairs of primes differ by N ? Thanks, Dan