wonders why create different 3D models, if you can create one that can be customized?
From the early days of TMP, we have always wanted to also do something on video, but… well, we tried, but we weren't very good at it! 
Recently, however, we realized that we could use low-budget AI for certain types of video projects, and it would look a lot better than getting our cellphones out, and cost a lot less than going into a studio. So let me step you through our first video project.
Editor Katie volunteered to be the guinea pig, as she has a keen interest in AI and videography.
The first step was to take her selected self-portrait, and use AI to make a few changes – hair, makeup and outfit, just as you would do in a studio. Here was the result:
Note that we're using what is called an AI Editor, designed to only change what you want to change in a picture. That's really Katie, just with new hair and a new dress. Her head may be out of proportion to her body, that's due to the angle of the original photo.
Remember, I said low-budget AI? We're going to make three 10-second videos, and string them together to make a 30-second video. The trick is to begin and end each video with the same image (except the ending), so they will (more or less) link up.
This means writing scripts in what you hope will fit in 10-second chunks. As a concept, we decided to do a light-hearted video about miniature wargamers being great lovers! (Yes, it's supposed to be amusing.)
Note that the voice is AI-generated. The AI seems to vary the voice based on the picture, and I have not succeeded in being able to modify the voice, no matter what I tell the AI. This is what it thinks Katie should sound like!
None of the acting – the gestures, the expressions – were specified. The AI 'came up' with all that. Similarly, the special effects were serendipity.
The dice roll at the end of the video was so nice, we tried to get the same effect in other videos – but the AI guardrails kicked in. Apparently rolling dice is sometimes against the rules!
What would we do differently next time?
The AI added some background music, but it doesn't match up because the video is three separate pieces – in future, I should specify 'no music'.
The voice is slightly out of synch in the first 10 seconds. With a bigger budget (LOL), I would have redone that segment.
We did ask for a laugh in the first segment. To me, it sounds a bit like Yoda. But I liked everything else, so I kept it. There isn't a way, currently, at our budget levels, to use AI to edit a video – it's all or nothing.
I also realized that the photo I used was the wrong aspect ratio (not tall enough). That's why you may see black strips at the top and bottom.
So that's how Editor Katie made her first TMP video!
I've also posted the video as a reel on the TMP Facebook page, where it has been one of our top videos. It is also reaching a younger demographic than our usual audience.