Did the well-to-do ladies of the manor in the 18th and 19th centuries like to soak their beautiful, expensive dresses? Well the answer is a resounding "yes!" Not only that, but their friends visited as well to join in the fun! The dialogue in these 4 video clips of 8 seconds each was entirely generated by Veo 3.1; I just supplied the broad topics of conversation and then the AI did the rest.

There are many pitfalls to generating AI videos featuring multiple people, so it was a frustrating exercise in some ways but a delight in others. For example, I love seeing the women in the clips interacting with each other, even when they're not talking, albeit briefly of course. And why just get one gorgeous dress wet at a time when you can do it with 4, 5 or 6?!

To name but a few of the pitfalls: multiple people will say the same dialogue at the same time, sentences can be heard but no-one speaks them, objects appear and/or disappear, there can be issues with collision detection and so on. Consequently, it took many video credits to end up with these 4 clips, but I'm pleased with the result, although I'm sure there will be a few oddities to be found if you look hard enough...