Imagine telling a computer, in plain English, what you want it to do in a science lab. No more complex coding, just simple instructions. That's the exciting potential of Large Language Models (LLMs) like ChatGPT in materials science. Researchers at the National Renewable Energy Laboratory (NREL) have shown how LLMs can generate control software for scientific instruments, starting with a common electrical measurement device called a Keithley 2400 Source Measure Unit (SMU). Through a conversational back-and-forth with ChatGPT-4, they built a Python-based control module and even a user-friendly graphical interface, all without writing extensive code themselves. The LLM learned the instrument's communication protocol (SCPI) and generated code for tasks like setting measurement parameters, controlling the instrument's panel, and executing complex current-voltage sweeps. It's like having an AI assistant that translates your scientific needs into machine instructions. But the NREL team went further. They combined this AI-generated control software with a high-performance optimization algorithm. This combo allowed them to automatically extract key device parameters from current-voltage data, providing insights into semiconductor behavior. They applied this tool to a novel high-power, high-temperature electronic device, demonstrating the practical value of this approach. This research shows how LLMs can democratize access to automated experimentation. Researchers without coding expertise can now control instruments and analyze data more efficiently, opening up new avenues for discovery. While this study focused on a specific instrument, the implications are far-reaching. LLMs could be used to automate a wide range of lab equipment, ushering in an era of AI-driven scientific exploration. Imagine the possibilities: faster experiments, automated data analysis, and a truly interconnected research environment. The future of science might be closer than we think, and it might just be written by AI.