I've observed something interesting: my toddler can quickly and somewhat effortlessly solve a simple physical puzzle, yet when I try to prompt large language models (LLMs) to guide me clearly through the solution, they struggle.<p>Even with clear photographs and detailed instructions, LLMs tend to give general advice or indirect methods rather than precise, actionable steps that clearly demonstrate correctness.<p>Have you tried something similar? Have you successfully "prompt-engineered" an LLM into giving clear, precise, step-by-step solutions for physical puzzles? If yes, what's your approach?