Wow! The expression of attitude, of I guess thoughtfulness, apprehensiveness, of something, on these, i suppose, robot's faces--it's so subtle, but still definite, and I can't really pinpoint where it comes from: a combination of scene and surroundings, of posture, eye direction, and facial composition? I don't know, but it is really striking.
And I guess I don't know where these figures themselves come from. I'm wondering, now, how they were created. Were they somehow drawn, or borrowed, or modified? This might be a complex technical question, but it is something I just don't know. Not asking for a full how-is-it-made explanation, but maybe, if you feel like it, a quick idiot's guide primer. The result, however, at points, is beautiful.
And, parenthetically, I've written an adaptation of a SF novel that has robots, androids, simulacra, whatever, in it, and there is some question about the degree of their sentience. They look and act human, but are somehow identifiable as not, and I haven't worked out exactly how. It is a decision to be made, i think, down the line, to be brainstormed, and one idea is that they might have something like an engraved "R" on their foreheads (maybe mandated by law) or maybe some physical functional difference.
For example, I have a guy, a repairman, encountering a secretary and asking her if she is having any difficulties. She looks up and says: I'm a human being. He sees and, embarrassed, says: Of course, I'm sorry. Anyway, this video gives me some ideas.
And, first impression, depending on the context, on what is going on behind this trailer, it might like music a little less bombastic, less overtly dramatic. Not criticizing, i think it's good, but just speculating.
Anyway, this one kind of knocks me out. The variety, the extravagance, the complexity, the, I guess, coolness of these scenes, with the watchful presence of their varied inhabitants--I start to fill in a story. Revolt!
Thanks, glad you liked this one.
To try and provide at least partial answers to your questions (the full answers would be way too long to get into)
The characters are not borrowed or derived directly from any pre existing source. They can't be. Save point will feature a populated universe, as you can already see from these sub reels, and it just wouldn't have ever worked out to try and source them. Too many characters, legal problems, expense, etc. There's tons of ways for me to get characters, but the logistics of this specific product make that impractical.
Here's where it gets interesting. Save Point characters are not exactly made, but rather grown. I'll try to make this as simple as possible. It's not going to be 100% accurate, due to the resolution limits of layman's terms. Basically, every character and object, every spaceship and planet, is generated from scratch every time you see it. Any visual idea you see is driven from a "Genetic" which I hand craft. A Gene is basically a knowledge base, that can be provided to brains called models, or packaged into memory units native to the models.
Genes are constantly "spliced", meaning that a memory for design x can be seamlessly hybrid with a memory of design y. In the example of the main character "Max" I've basically trained a memory on footage and images, and then tried different mixtures until I got the result I wanted, basically 30% Clint Eastwood, 30% Christian Bale, 15% Anthony Jesselnick, 15% Brandon Beemer, and 10% random photos from men's watch advertisements. Nothing is copied, I just train the memory unit to answer what a face should look like, and I show it pictures and videos of faces, and it forms an impression or "memory" of what a face looks like. If I didn't segment that particular brain when dropping it into the pipeline, the entire crowd would have faces derived from that units limited experience with what faces look like. Thus segmentation comes in, where multiple AI brains and modular memory units are used in combination to draw a single scene. Hey robot, let's spend 30 hours talking about, and looking at pictures of a white future city. Put that brain in a jar and label it. Hey Robot, this is what a servant android would look like, memorize this genetic. Put that brain in a jar and label it. At the pipeline stage, I'd plug in jar 1 and jar 2, set up camera angles, dof, and direct the scene. Hey pipeline, let's shoot a 5 second crane shot moving slowly in on a servant android looking apprehensive in a white future city. Plug in another jar for a memory of how crowds move in theme parks. Pipeline, add crowds.
I think it's worth noting that this is exactly what human artists do when they create an "original character". In example, I'd call Han Solo an original character, but what he really can be boiled down to is Lucas mixing together various iconography he'd seen in other films, or elsewhere. Gun belt, bandolier, cocky attitude, worn dusty leather jacket. All this stuff comes from the westerns Lucas grew up on. It got all jumbled up in his head from a 1000 impressions, and became something new, which we now perceive as an original character. I'm doing the same thing, just faster and more often. I need to be able to scale to infinite size with no cost, so that's the solution I engineered. Just to be clear, the AI components I create do not copy any of what I show them, they just get impressions of what X is like, then I task them with creating something new, once a mentality has been formed from those impressions.
As far as what robots are like in the Save Point story, I'd say that they are incredibly diverse, and don't fit neatly into one category. In this video above, I wanted to show off what I think is the most interesting type of robot, the ones that could almost pass for human. Basically, the facial expressiveness I've been able to achieve this way opens the door for many types of stories that simply wouldn't work with metal faced robots. I'll never try to answer the question or AI sentience in the story, at least not through robots. I think the tension can thrive within a dangerous ambiguity, and having the android variants be emotionally expressive allows more sophisticated storytelling. Anyway, I thought this clip pretty much sold the idea of robots with nuanced facial expressions.
This storyline for example is exactly what I'm describing. This would be one of the "episodic" adventures within Save Point, where Max visits a planet with a caste system of giants, humans, and their highly evolved robot servants. This glimpse takes place at a pivotal moment when a terrorist network of ideologist androids assassinates a political leader, thus signaling the beginning of revolt and revolution. Lot's of interesting options for a choice based story that can have a lot of action, high stakes, and thought provoking moral quandaries about who is right and wrong in a very grey area where key information (Can androids feel? Can you kill something that was never alive? What constitutes being alive?, etc.) is unavailable.