Intro to TLA+ for the LLM Era: Prompt Your Way to Victory
Summary
The article introduces using TLA+ in the era of large language models, arguing that frontier LLMs can generate TLA+ specs while humans define correctness and the meaning of 'correctness' using temporal logic. It uses a beans puzzle to illustrate initial states, state transitions, and model-checking concepts, and demonstrates prompting an LLM (e.g., Claude) to produce a TLA+ specification.