ElevenLabs V4 AI voice and Seedance 2.5 AI video update

I’ve been working with a few AI video and voice models to create professional training video for bus mechanical work. We’ve had some excellent results using ChatGPT image and Grok Imagine for source imagery, and Seedance 2.5 and 2.5 Turbo for video creation, then ElevenLabs V4 for extremely natural voice production.

The process of creating the video itself is slow and requires a fair bit of management and babysitting. Often the source image creation needs multiple attempts to get the details right, and then a similar problem with Seedance, but the difference is that Seedance takes 5-10 minutes to create one 5-10 second clip, which often has to be recreated.

With my 10-year background in photography and video production, one thing that’s interesting is that it’s actually faster to film and edit a video as a human, with humans, rather than creating everything from scratch using AI. But obviously the advantage with AI is that we can literally do anything… with enough time, prompting and vision.