Simulating Opinion Dynamics with Networks of LLM-based Agents