Like with all generative AI this comes down to how specific you can get, and how consistent you can be. If you can art direct each element of the frame, down to the design of the individual props and scenic items, and have those items remain consistent from shot to shot. Then do the same with lighting, actors, camera characteristics (eg lens, focus, position in the scene, framing), etc, etc, etc, then maybe you’ve got a chance of making a ‘high budget movie from the early 2010s’. But I haven’t seen any generative ai that comes even close to this level of control or consistency… They’re closer to slot machines than anything consistent…
Yup, you can notice some issues even in their picked example. e.g. the prompt for the video of the painting woman says "there is a bear cub at her feet" and it quite clearly is not "at her feet" in the video.
Hi I'm Max, this is my project. As you say the text sizes on the Map screen are inconsistent in the movie. I did wonder if this could be to do with image/lens distortion, either from the camera or the glass on the CRT.
I considered drawing them out pixel by pixel to match the movie exactly but in the end I just settled for erasing the cap on the J in San Jose, so it would match the movie more closely.
