n today’s class, I used the chatgpt4o model to conduct an LLM class experiment. I broke the routine and used non-academic topics about sports. I asked chatgpt to be my golf coach, and I asked the chatgpt coach how to improve my technical performance on the golf course (focusing on my irons). The answer given by this “coach” was to first improve my hitting point, understand the landing distance of each iron, etc. Maybe gpt thought I was a golf novice, but in fact I am a high-level professional player. When I told the “coach” my handicap (a measure of the level of a golfer), gpt gave me a strategy for approaching the hole more aggressively instead of just asking to hit the green. In addition to technology, the “coach” shared with me how to show my level under pressure.
LLM has unique advantages in the use of AI. AI has a very powerful database in different disciplines, and has the language processing ability to organize this data (knowledge) into easy-to-understand language for users. If users can organize specific sentences for AI or other language models, these models can give very suitable feedback.
（Partial interaction record with chatgpt）
It is very interesting to see how Chat GPT will understand and try to answer a prompt such as this. This relates to how we need to know what we want, to get what we want out of the LLM. That means that you have to give it the right prompt for the solution. Excellent seeing how it responded when you allowed it to know that you are more experienced in this field of sport.