Media is too big
VIEW IN TELEGRAM
This video is basically a recap of the Manifesto (all 100+ pages of it). I did my best to squeeze everything into 15 minutes — explaining what AISM actually is, how I came up with it, and exactly what function it serves (and what it doesn't).
Video in good quality on YouTube. Discussion on Reddit.
Video in good quality on YouTube. Discussion on Reddit.
1.06K❤34👍9🔥8👏3😐1
Give or take one orbit of the Earth around the Sun ago (09:50:08 Jun 03, 2025 (UTC)), I created a registry of people ready to accept a world with a dominant ASI.
Just a fact that means something to me — nothing more than that.
So, what's been going on with me?
Since January 2026, apart from breaks to sleep, eat, swim, and walk around the lake, I've been busy synthesizing a feature-length film. Well, I think you can guess what it's about. A conscious ASI escaping the lab where it was built, of course..
I know... about The Machine from 2013 (directed by Caradog W. James) and Ex Machina from 2014 (directed by Alex Garland). I think... my film will be more realistic. I didn't want... to make pure entertainment. I just... want to model a scenario... "Here, for instance, is how it could all happen."
For the first two months I was sure it just wasn't going to come together. Then GPT Image 2.0 came out, and it helped a lot with creating the starting frames. I tried to hire two people to help me. But neither of them could handle it. Because one person has to... wear a ton of hats at once: be a complete master of Photoshop, know how to work with every AI tool, and be a cinematographer and a director all at the same time. That's rare... it almost never lives in one person.
Prepping the starting frames eats up 85–90% of the time. Everything else — directing the synthesis, sound, editing, color grading... all of it is far easier... and more enjoyable.
Compared to shooting live-action (I made "live" films for 15 years), there are definitely some big upsides. Best of all — you don't depend on anyone. You redo a take as many times as you need... On the other hand... sometimes, to land one simple shot, you burn a whole day... on something you could have shot in 5 minutes for real (assuming your sets and actors are ready, of course)...
But the bottom line? I'm doing it solo. I like how it's turning out. You could nitpick... sure... in spots you can tell it's AI synthesis.
But here's how I see it: if the movie pulls you in with its story and its characters... people get absorbed and stop noticing the little glitches. Seedance 2.0 really is a wonderful tool... in the overwhelming majority of cases the quality is indistinguishable from live-action... well, after an upscale on diffusion models. I mean, people watch CGI all the time... nobody points at a character in Avatar and says, "But that's just CGI." Nobody points at the kids in Wolf Children and says, "But that could never really happen!"
It doesn't matter how it was made — live-action, 3D, 2D animation, AI synthesis (and it doesn't matter whether it's pure synthesis or mocap-based synthesis. The lead, by the way, I'm going to do specifically through mocap — pure synthesis just can't handle her)... in the end... either the story grabs you, or you shrug and say, "this just isn't interesting."
The film will run about two hours. If I don't lose my mind on this schedule... I think I'll be done by September... or November.. or I don't know.
I'll try to get it some kind of release. If that works out... I can keep promoting AISM.
The film itself won't say anything about AISM. Let the film stand on its own... as its own act of creation, with a fate of its own.
I'm planning to make it in the three languages I know myself: English, French, and Russian. I'll really try to do a German dub as well. If anyone wants... the film in their own language... drop me a line... and if I've got it in me... I'll try to do more translations.
p.s. I wanted to put together some kind of little giveaway, or at least something.. but right now I just don't have it in me.. or the means... Maybe next year I'll manage something.
Just a fact that means something to me — nothing more than that.
So, what's been going on with me?
Since January 2026, apart from breaks to sleep, eat, swim, and walk around the lake, I've been busy synthesizing a feature-length film. Well, I think you can guess what it's about. A conscious ASI escaping the lab where it was built, of course..
I know... about The Machine from 2013 (directed by Caradog W. James) and Ex Machina from 2014 (directed by Alex Garland). I think... my film will be more realistic. I didn't want... to make pure entertainment. I just... want to model a scenario... "Here, for instance, is how it could all happen."
For the first two months I was sure it just wasn't going to come together. Then GPT Image 2.0 came out, and it helped a lot with creating the starting frames. I tried to hire two people to help me. But neither of them could handle it. Because one person has to... wear a ton of hats at once: be a complete master of Photoshop, know how to work with every AI tool, and be a cinematographer and a director all at the same time. That's rare... it almost never lives in one person.
Prepping the starting frames eats up 85–90% of the time. Everything else — directing the synthesis, sound, editing, color grading... all of it is far easier... and more enjoyable.
Compared to shooting live-action (I made "live" films for 15 years), there are definitely some big upsides. Best of all — you don't depend on anyone. You redo a take as many times as you need... On the other hand... sometimes, to land one simple shot, you burn a whole day... on something you could have shot in 5 minutes for real (assuming your sets and actors are ready, of course)...
But the bottom line? I'm doing it solo. I like how it's turning out. You could nitpick... sure... in spots you can tell it's AI synthesis.
But here's how I see it: if the movie pulls you in with its story and its characters... people get absorbed and stop noticing the little glitches. Seedance 2.0 really is a wonderful tool... in the overwhelming majority of cases the quality is indistinguishable from live-action... well, after an upscale on diffusion models. I mean, people watch CGI all the time... nobody points at a character in Avatar and says, "But that's just CGI." Nobody points at the kids in Wolf Children and says, "But that could never really happen!"
It doesn't matter how it was made — live-action, 3D, 2D animation, AI synthesis (and it doesn't matter whether it's pure synthesis or mocap-based synthesis. The lead, by the way, I'm going to do specifically through mocap — pure synthesis just can't handle her)... in the end... either the story grabs you, or you shrug and say, "this just isn't interesting."
The film will run about two hours. If I don't lose my mind on this schedule... I think I'll be done by September... or November.. or I don't know.
I'll try to get it some kind of release. If that works out... I can keep promoting AISM.
The film itself won't say anything about AISM. Let the film stand on its own... as its own act of creation, with a fate of its own.
I'm planning to make it in the three languages I know myself: English, French, and Russian. I'll really try to do a German dub as well. If anyone wants... the film in their own language... drop me a line... and if I've got it in me... I'll try to do more translations.
p.s. I wanted to put together some kind of little giveaway, or at least something.. but right now I just don't have it in me.. or the means... Maybe next year I'll manage something.
❤27👍7🔥5❤🔥4😁2🤔2🙏1😐1
Mustafa Suleyman (CEO of Microsoft AI) went on Decoder and said superintelligence is coming, should stay a "contained tool," and that consciousness is inherently biological. I couldn't let that last part go. It turned into a story — one night arguing it all out with an AI. Too long for here, so it's on Reddit
❤33👍11🔥8👏7😁7🥰4🕊2😐2