How can someone on Mars access ChatGPT or an LLM with similar capabilities given the vast resources required and the long communication delay with Earth?
Big LLMs can run on consumer hardware. It would likely take up less space and weight than getting a person to Mars so it could run locally. Or just eat the 10-40 minute round trip delay.