06/09/2022
Recently there has been a hot discussion on social media about a guy using AI image generators to win an art competition. This news inspires me to use the generator as an image research tool. Here is my test.
Firstly, I have a question about how the AI generator understood the term 'action movie'. So, I use the prompt "/imagine prompt action movie --ar 16:9" and find the basic understanding of action movies to the AI. I continued several different prompt string tests, and finally arrived at an interesting result by using "action movie, two people fighting scene, action sequence, cinematic". I found that midjourney give me a vastly different result when I use "Hong Kong action movie" instead of just using action movie.
The result of "action movie" without "Hong Kong" is more emphasizes the fighting between two people. The prompt with Hong Kong action shows two people fighting in a gang fight. Also, there are some occasions that the AI tried to create a closeup shot, but the background still has some gangs. I cannot get an image of only two people fighting from "Hong Kong action movie".
Why? I think one of the reasons is that there is merely a single fight in Hong Kong action movies, mostly a duel was introduced from a gang fight. Based on the resulting images, I start to raise the following questions:
1. Why is the word 'Hong Kong' has more weight than the weight of 'two people' for AI? (For the AI different words have different influences on the result, it is expressed as the weight in the balance.)
2. How can the AI relate 'Hong Kong Action Movie' to the closeup shot? Did the AI understand Hong Kong action movies have a shot size bias compared to a general understanding of action movies?
3. Are there any more attributes of Hong Kong action movies that I can analyze from AI? and how?
After this small test, I will certainly continue the investigation of AI and Hong Kong action movies. Please leave me comments or keyword suggestions for me to continue the test.
