AI-Enhanced Educational Video Production

I dont want to waste your time, or mine. So please read the brief from start to end. It’s probably going to take you about 30 minutes, so you need to be willing to invest this time in understanding the brief. I dont want to repeat myself, to tell you to ‘read the brief’, so please do not contact me until you review all of this thoroughly. This could be a big and long term project. But I dont know if AI is at the point yet. Or if there is anyone out there with the skills and ability to grasp the content sufficiently. Firstly, I need to take you through the history of this project. 10 years ago I produced an extensive lecture series on all my Gestalt therapy training. 82 hours, 34 topics. Later I created a 3 minute video on each topic. So 68 videos in total. You can see all of that in my LearnGestalt site. Please view this before proceeding. Ok, having done that, you will see my style, and what I am doing - delivering professional teaching on a complex psychological topics, theories and practices. I also have an introduction, which has proved very popular - go to Youtube and type in: Gestalt in a Nutshell Steve Vinay Gunther Ok, you get a sense of what I am doing. So, 6 years ago I got the idea to get explainer videos done for all the lectures. I spent 18 months, a lot of discussions, samples, I had the idea of whiteboard drawing, but soon realised the moving hand was actually distracting, so later ditched that idea. I put up a video of an introduction to Field theory: Attachment: 1-1. Field theory 11 mins And the transcript for it: Attachment: 1-2. Field Theory transcript complete - a section is highlighted, which is an excerpt I made of a few minutes. I took the excerpt and put myself in the corner: Here was a video I started with, putting myself in the bottom corner: Attachment: 1-3. Field Theory - Excerpt trial video Now you can see three samples of what people did to storyboard that and make into a video: These is the transcript for that section: Attachment: 2-1. Proposal A Attachment: 2-2. Proposal B Attachment: 2-3. Proposal C I chose the last one - C. I moved ahead with them. Here are some storyboards they created: Attachment: 3-2. Storyboard - Field Theory Attachment: 3-3. Storyboard explore Attachment: 3-4. Gestalt Overview storyboard I really liked the clean style, the subtle graphics, the simple colours, the overall aesthetics. I planned on doing all 82 hours/34 lectures like this. As you can see, we got started, then unfortunately they completely ripped me off, and I gave up. I did talk with some graphic artists who came up with another, more organic proposal. Here is the overview document: Attachment: 4-1. Organic Proposal Here is a video they did: Attachment: 4-2. Organic video They experimented with some looks: Attachment: 4-3. Exploratory video 1 Attachment: 4-4. Exploratory video 2 Attachment: 4-5. Exploratory video 3 Attachment: 4-6. Exploratory video 4 Attachment: 4-7. Exploratory video 5 This was very promising, but it seemed too hard, and although I really liked their aesthetics, their ideas, it was hard to see something that would be a clear expression of each concept, plus the size of the project with this studio approach would just be too big. Enter AI the world of AI. Maybe AI videos could make this possible? So I am exploring this possibility. I am interested whether you have the design capacity, AI skills, aesthetic sense, and grasp of the content enough to come up with something new. If you do, this could be an ongoing collaboration. But firstly, I need to be very very clear what I DO NOT want: ** I am not having *any* flashing text transcription of my voice. None. **I do not want *any* human replica figures - the hyper real style of most AI video. I also dont want the sci-fi version of hyper real either. **No voice over required - the lecture voice itself will be the sound used. **I do not want another figure speaking my words. Use a smaller excerpt of me speaking the words. **Panning between still images could be ok, but they would have to be of an equal quality to the #3 storyboard drawings. Nothing less is acceptable. Same with cartoon type figures - nothing less than the subtly and quality of these storyboard drawings would be acceptable. **Each image or movie sequence has to relate directly to what is being talked about - no extraneous elements. Very clean, direct, focused on extrapolating the concept. **The style needs to have a consistency **I do not want any stock images - these always look like stock images, and I certainly dont want a pastiche of them **My total preference is ORGANIC, symbolic images. To this end, in my amateur way, I have explored this. I took the Overview lecture Attachment: 10-1. Youtube link to Overview lecture Fed the transcript: Attachment: 10-2. Overview transcript To DeepSeek, gave it this instruction: “Generate Sora prompts based on the attached lecture, using symbolic images or figures, not human replicas” The result: Attachment: 10-4. Overview lecture Sora Prompts Generation - DeepSeek I took Prompt5, and fed it into SeeDance2.0 The result was this: Attachment: 10-5. AI movie from prompt 5 I then combined the excerpt and the AI movie thus: Attachment: 10-6. Split Overview excerpt and AI movie prompt 5 Now, this is my first exploration of this. I do not have expertise. You do. So I expect you can do a lot better at giving AI instructions for prompts, getting more refined prompts, and then getting more refined videos. You can see this is organic in the way I am wanting. I would be interested in the cartoon version of the previous Storyboard guy, and maybe it’s possible to train up AI to recreate this. That would be a different approach. I dont have that expertise or experience. Do you? At present, with this organic approach, my idea is that each concept would have an accompanying video illustrating it. The transition between each video might be a few moments (or more) of a ‘primordial soup’ - a consistent very slowly moving ai video…’out of which’ each concept video arises and ‘into which’ it dissolves. A bit like the old Lava lamps. Perhaps in some way like the ~4-2. Organic video~ example, though a little softer. 
So this is exploratory. If it works out, then I might make it into an extended project. I certainly am not interested in garish productions. They must have a sense of aesthetics, and be suitable for the professional setting and content. And as with my efforts 6 years ago, I realise that you cannot do this unless you have some kind of grasp of the concepts. 
 Another approach is here: Attachment: 11.1 Another approach This little cartoon Pi character is cute without being cringeworthy. The story telling is very clear. BUT - it’s maths and science, and that is a lot easier to visually demonstrate than the complex psychological concepts I am talking about. I also guess this is not strictly AI, but I dont see why AI couldnt be directed towards something like this. However, it still needs to be directed. And obviously I wouldn’t have a PI character…so I am not sure how useful this example is in the end. There are the videos of Sabine Hossenfelder on YouTube, but I dont want this style. The images are very much stock. She is super fascinating and topical, so her content overshadows the cheap images and clips. But I just do not want anything that looks like this. It’s very pastiche, and very much looks like a quick hack job. She gets away with it, but I dont want to go down that road. — 
I am not a large enterprise, despite my solid website, so I dont have a huge budget. But I am willing to try something out, and if we can get something that works, I would start getting this applied to each video, probably starting with the 3 minute ones. So far, in all the responses I have received, I have not seen anyone provide me with any previous samples of their work that even comes close. So I am not going to look at what you have done previously. Dont send to me, it’s wasting our time. If you are interested in this project, you need to supply me with an initial demonstration of how you would fulfil my requirements. You can download the Overview video yourself. It’s too big to include here. I have provided with you the transcript, the timings, and my version of AI prompts. But hey, I can do all that myself. I am looking for your expertise to deliver something a bit more sophisticated, that shows me you can do more than I can do. So you can pull out a few minutes from that video and have a go at illustrating it with AI. It has to be more sophisticated than what I can do, in my attached example, with my very simple first pass. I am not going to pay for this sample, so please dont ask me. This is an investment of your time, to demonstrate to me your ability to do this. I am happy to give you feedback, and keep working with you to refine what you do. If this forms into something that I am confident about, then indeed, we can start launching the process formally. Please do not contact me until you have read and understood all this. You can ask me further questions then, but ultimately I want to see what you can do. ’Show dont tell’.

Proje No: 40277155