In an interview, Meta’s chief executive took aim at Anthropic and OpenAI, which have pushed to tightly control A.I. development, and said he supported “more openness.”