|
Suboceanic Reimagined by Tomasz Dobrowolski
[nfo]
|
||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
|||||||||||||
|
popularity : 54% |
|||||||||||||
alltime top: #85659 |
|
|||||||||||||
|
||||||||||||||
added on the 2026-02-22 19:05:53 by tomkh  |
||||||||||||||
popularity helper
comments
Well, shouldn't it be 64K?
added on the 2026-02-22 19:55:54 by fizzer 
fizzer: it would be nice to fit a video model with this level of fidelity under 64k, but then I would surely release it in "demo/intro" category, not in "Wild with AI".
FIIIIIIIRST
It's pretty amazing what these statistical models can do nowadays. I sort of wanted to give +1 because it was not bugging too much (atleast watching it once) and I sorta understand the excitement as an old procgen enthusiast.
..but. It's probably fine as a wild, but I have to say that I'm disappointed that this was just using those 3rdparty models. So it's not your own work, except resampling the good shots and then combining them on the video. I would loved to see you distill the models to produce this kind of quality in say 10 megs or so.
..but. It's probably fine as a wild, but I have to say that I'm disappointed that this was just using those 3rdparty models. So it's not your own work, except resampling the good shots and then combining them on the video. I would loved to see you distill the models to produce this kind of quality in say 10 megs or so.
It's certainly your right to do it.
But man, when I saw the A/B pics you showed a few days ago (wherever I saw that), of screens from this shown against screens from the original prod.. I thought this was an absolute tour de force of showing how your skills, plus the tech, had advanced in the intervening years.
So, now I'm sad. The realisation of what was actually going on was one the sadder collisions I've experienced of Good Old Stiff Upper-Lip Scening, against what someone else described as stochastic echoes in idea-space.
I don't think we've met, and somehow I don't like to thumb down someone I haven't looked in the eye, but this has made me feel ick.
But man, when I saw the A/B pics you showed a few days ago (wherever I saw that), of screens from this shown against screens from the original prod.. I thought this was an absolute tour de force of showing how your skills, plus the tech, had advanced in the intervening years.
So, now I'm sad. The realisation of what was actually going on was one the sadder collisions I've experienced of Good Old Stiff Upper-Lip Scening, against what someone else described as stochastic echoes in idea-space.
I don't think we've met, and somehow I don't like to thumb down someone I haven't looked in the eye, but this has made me feel ick.
Fell: this is indeed not a demonstration of my skills, but a "performance art" of sort. I explicitly labelled everything. Also from the first seconds there is a full disclaimer that I did "next to nothing" over here. Sorry, if this was confusing in any way.
It's a commentary on modern times.
Please judge it appropriately.
It's a commentary on modern times.
Please judge it appropriately.
So, I’m wondering. Did you enjoy making this as much as you enjoyed making the original? Or even more perhaps? Also, how is it in the ‘sense of accomplishment’ department? The original vs. remake? I’m genuinely interested. Thanks.
4gentE: no sense of accomplishment whatsoever. Again, "I did next to nothing". But as a viewer, I actually do like the outcome ;)
Thumb up for having the balls to submit an AI-prod :)
Fun to watch, curious about the whole pipeline, if you care to document it... :)
Fun to watch, curious about the whole pipeline, if you care to document it... :)
Should have been called »suboceAnIc«.
Interesting concept. Did you used single frames for this or was the AI-version interpreted from the whole video at once?
Interesting concept. Did you used single frames for this or was the AI-version interpreted from the whole video at once?
great that you finally made a demo about all that AI discussion
ah no wait, it's just a low effort video with more "social commentary" trolling
good job, well done, keep at it!
ah no wait, it's just a low effort video with more "social commentary" trolling
good job, well done, keep at it!
Glöp.
this looks like burned lifetime.
for me it just doesn't matter if it wasn't done by you to some extent. but as always it's fascinating what it can do --- still it's not about you. can't even thumb this down or up.
for me it just doesn't matter if it wasn't done by you to some extent. but as always it's fascinating what it can do --- still it's not about you. can't even thumb this down or up.
You do get an oink point for sparking discussion. But if the achievement here is converting a 64kb handcrafted work of art into 116MB of instant slop, I'm not feeling the excitement about it. I personally prefer the 64kb handcrafted works of art.
There are probably many ways to interpret the message here, but I was left with the thought "if you don't feel like making any effort, here's what you can do without effort these days". But that just doesn't feel like an inspiring hobby/scene for me personally. I guess I like effort.
There are probably many ways to interpret the message here, but I was left with the thought "if you don't feel like making any effort, here's what you can do without effort these days". But that just doesn't feel like an inspiring hobby/scene for me personally. I guess I like effort.
I think this was an interesting exploration on turning old prod to something new and my personal favourite on the wild compo.
The reason that I asked why it's not 64K is that the original was 64K, so to make a point about the effectiveness of AI you should have used AI to remake it in 64K so then we can make a comparison. Otherwise, what's the point?
psenough: try going out of your comfort zone once in a while... it’s more fun than playing it safe for the conservatives!
fra, gaspode: thanks! At least someone gets it.
My pipeline was nothing special for an AI production. I don't want to mention specific products here, not to be accused of promoting them, but it's all off-the-shelf commercial models. The transformation from original low-res images was done by LMM. I kept everything in context for consistent look. I did experiment with video2video + references, but it didn't work so well (or I gave up too fast), so it's all image2video, sometimes with both start and end frame specified.
fra, gaspode: thanks! At least someone gets it.
My pipeline was nothing special for an AI production. I don't want to mention specific products here, not to be accused of promoting them, but it's all off-the-shelf commercial models. The transformation from original low-res images was done by LMM. I kept everything in context for consistent look. I did experiment with video2video + references, but it didn't work so well (or I gave up too fast), so it's all image2video, sometimes with both start and end frame specified.
After writing my comment, I realised it's not obvious in pouet that this was an "AI allowed/encouraged" compo ("Wild with AI" it was called I think).
So in that context my comment wasn't entirely fair and I should say this was more interesting than generic GenAI output.
So in that context my comment wasn't entirely fair and I should say this was more interesting than generic GenAI output.
As a commentary on modern times: I'm not sure what I'm supposed to take away from this.
As a wild demo: No.
As a wild demo: No.
sagacity: it's up to you what you take out of this. Or do you prefer being told what to think? ;)
Nice AI work, but sad that's not an humans' creation.
Compo was titled "Wild with AI", and production slide clearly stated how this was done. I enjoyed this, and helped me discover the original. For those reasons, it deserves a thumb from me.
What Mysdee said above. Compo name was Wild with AI. Tomkh went with it and that's no crime. Besides it was prettyyyyyes.
Mysdee, mikron, hepphep: thank you! I'm glad you enjoyed it.
you sure THIS is your commentary on modern times... not your 100s of posts in the BBS about AI?
Maali: at least I don't hide my AI creations - maybe you should release some of yours too. I'm honestly disappointed by your thumb down!
Are you disappointed by mine?
fizzer: I fully understand the aversion to AI, but then why you even bother watching this. Just enjoy other compos where there is no AI. Live and let live!
Well, why did you release it if don't want me to express an opinion about it?
fizzer: I have absolutely no problem with you expressing an opinion! It's just not an opinion about this particular prod so far.
To answer your comment about 64kb limitation - I completely agree, it would be much more impressive if AI could "spit out" code, music etc.. with decent quality that compiles into 64kb executable, but it's not yet able to do that (last time I checked). Also a remake doesn't necessarily has to be in the same category as the original. But thank you for sharing your thoughts on this.
To answer your comment about 64kb limitation - I completely agree, it would be much more impressive if AI could "spit out" code, music etc.. with decent quality that compiles into 64kb executable, but it's not yet able to do that (last time I checked). Also a remake doesn't necessarily has to be in the same category as the original. But thank you for sharing your thoughts on this.
nope
First you say "why did you watch it" and then you say "you haven't said anything about it".
I wish I could give an extra strong piggie for how completely devoid of any feeling this has left me.
Cool reimagination of your work!
Mermaids are supposed to be beautiful... This one looks like she crawled out of hell.
The original 64K intro is much better!
I'll close my eyes and give it a thumbs up anyway. :]
The original 64K intro is much better!
I'll close my eyes and give it a thumbs up anyway. :]
Dystopically soulless cry for attention
It looks like slop
darya: dystopian, yes! The line between human and machine starts to blur - we are slowly becoming one.
NR4: luckily, I'm not performing a colonoscopy here, but rather participating in the Wild with AI compo!
:)
Quote:
It's pretty amazing what these statistical models can do nowadays. I sort of wanted to give +1 because it was not bugging too much (atleast watching it once) and I sorta understand the excitement as an old procgen enthusiast.
..but. It's probably fine as a wild, but I have to say that I'm disappointed that this was just using those 3rdparty models. So it's not your own work, except resampling the good shots and then combining them on the video. I would loved to see you distill the models to produce this kind of quality in say 10 megs or so.
Would have been neat indeed to do it on a completely self-made/hosted model. As in, train exclusively on your own material, with no outside things.
Then again, even to some people, that's worthy of being thrown into the seven circles of Hell alone :/
I respect that people have different opinions about AI.
However, ganging up against an AI production submitted to a dedicated AI competition is crossing the line. It feels less like a critique and more like "be like us or be against us" mentality.
I know this doesn't represent everyone and it's likely just a few individuals, but it’s turning the scene into an unwelcoming place where newcomers (or frankly, everyone) have to constantly watch their back and filter their creativity to avoid harsh backlash.
However, ganging up against an AI production submitted to a dedicated AI competition is crossing the line. It feels less like a critique and more like "be like us or be against us" mentality.
I know this doesn't represent everyone and it's likely just a few individuals, but it’s turning the scene into an unwelcoming place where newcomers (or frankly, everyone) have to constantly watch their back and filter their creativity to avoid harsh backlash.
Just came here to clarify that this was not a dedicated AI competition. I can see how the compo title might be misleading, but a quick read of the rules should have made things clear. Or look at the other entires (in this or previous years) where you can clearly see that not every one involved AI.
It shouldn't be too hard to understand that many people are here (since decades, respect that everyone saw a lot) for creative content made by human. Again, to some extent we all use tools (Photoshop, a bit AI here and there, synthesizer presets, even Fairlight "reimagined" [mostly copied] a song by Vangelis/Socrates in Agenda Cycling Forth!) but AI and slop made all of this crossing a line so this content just isn't interesting anymore. That doesn't mean this becomes an "unwelcoming" place, this place just doesn't want to adapt to the slop everyone now has to watch in movies, tiktok and everywhere else. It doesn't even mean AI art isn't art or whatever definition you want to choose. As I said, this content just doesn't matter - and even this compo doesn't matter, if not people made it to a better part.
The demoscene values human craft and now fights against the move. Don't make us stand like ignorant idiots here please.
The demoscene values human craft and now fights against the move. Don't make us stand like ignorant idiots here please.
the more you fight against it, the more he comments back, which give his absolute slop of a release some attentions and he thinks he's winning. Dude does not get enough attention at his job in AI (just a wild guess…); wondering if he's the one he says he is, the one that seems to be living in Berlin (lots of not-so-nice leftists there), at Google maybe? (Lots of nice people there).
evilpaul: my point stands: it wasn't only dedicated to AI, but the rules were explicitly encouraging for full-AI generated entries. On top of that, the title of the competition was Wild with AI, which should make it crystal clear what to expect.
And the majority of entries in the compo were using AI to some extent. I wouldn't be surprised if, after seeing the backlash here, these authors will never dare to add their prods to pouet. And I don't blame them.
And the majority of entries in the compo were using AI to some extent. I wouldn't be surprised if, after seeing the backlash here, these authors will never dare to add their prods to pouet. And I don't blame them.
what evilpaul said... this was the 'let's sacrifice the wild compo so we can avoid (or easier justify disqualification) AI in all the other ones'-compo.
I find it curious that people can prompt something with no effort (author's assertion not mine), yet be so emotionally invested in it. Not the only prod I've seen this happen in.
very nice idea to let the AI reimagine an old prod. perfect fit according to compo rules and I was properly entertained.
Byproduct: I don't mind comments like "looks like slop", "not fun to watch" - those are absolutely fair. But what I find peculiar (not that I'm particularly surprised) is personal attacks and complete backlash against the very fact it's AI used in AI-allowed compo.
Interesting experiment. While the visuals are nice, the camera movements and the connection/pacing between the shots seem rather chaotic. I understand that this is an experiment in doing "next to nothing", but I am curious about what could be achieved if more human effort is put into these aspects.
It occurs to me that 10 kWh doesn't seem like a lot compared to what visuals like these would consume through traditional offline rendering.
I hope this wave doesn't split the scene in half. There should be room enough for all of us.
It occurs to me that 10 kWh doesn't seem like a lot compared to what visuals like these would consume through traditional offline rendering.
Quote:
I wouldn't be surprised if, after seeing the backlash here, these authors will never dare to add their prods to pouet. And I don't blame them.
I hope this wave doesn't split the scene in half. There should be room enough for all of us.
Such a split has happened already in other communities, so it's not impossible.
Quote:
if the achievement here is converting a 64kb handcrafted work of art into 116MB of instant slop, I'm not feeling the excitement about it. I personally prefer the 64kb handcrafted works of art.
Don't find the meaning of this, but whatever rocks your boat :)
Quote:
However, ganging up against an AI production submitted to a dedicated AI competition is crossing the line. It feels less like a critique and more like "be like us or be against us" mentality.
It's a website for personal opinions on prods. You are allowed to hate AI, even if it was an AI friendly compo.
wrighter: so you think it's ok to "hate"? Wow.
A difficult one.
On the one hand it's totally fine to do such kind of a "remix" of your own work,
I just wouldn't put it in a compo, though. But on the otherhand Assembly offered exactly this kind of compo. And although it's interesting to watch and compare, I find it difficult to give a thumb for writing prompts.
However, a well played piece into the ongoing discussions around A.I.
On the one hand it's totally fine to do such kind of a "remix" of your own work,
I just wouldn't put it in a compo, though. But on the otherhand Assembly offered exactly this kind of compo. And although it's interesting to watch and compare, I find it difficult to give a thumb for writing prompts.
However, a well played piece into the ongoing discussions around A.I.
You think that a community so deeply rooted in human artistry can just shrug away at other people's willingness to accept the complete absence of it as an achievement? Sorry, but you're mistaken and it's completely within our rights to state our opinion, just like it was yours to release something within the bounds of a compo.
hfr: it wasn’t a compo „exactly“ about that. The compo was called „Wild (with AI)“ - in brackets. Just a wild compo where they also allowed AI generated entries.
Actually, let's be clear, it would come from any person who did not know better, I'd be a tad more chill about it; but it comes from a particular individual who has repeatedly tried to stir the conversation about AI while at the same time victimising himself and trying to make others feel like they were the baddies.
Honestly, OP keeps saying he's apolitical but he's certainly using the same sort of technics used by Very Smart, Polite And Well Dressed Gentlemen.
Honestly, OP keeps saying he's apolitical but he's certainly using the same sort of technics used by Very Smart, Polite And Well Dressed Gentlemen.
i wouldn't mind a proper AI exploration video, but this is just low effort framed for more attention seeking on an already over-discussed topic, and that's what i dislike about this video.
AI can be used for cool things! like this very half-assed video just doing some basic image 2 video of an old entry i did. does not impress, just troll for attention and rinse and repeat of more senseless discussion around AI.
if your serious about "we can do cool things with AI" then fucking do cool things with it, this ain't it.
AI can be used for cool things! like this very half-assed video just doing some basic image 2 video of an old entry i did. does not impress, just troll for attention and rinse and repeat of more senseless discussion around AI.
if your serious about "we can do cool things with AI" then fucking do cool things with it, this ain't it.
darya: this community was never purely artistic. Please don't forget it was technical oriented community at the very beginning. It was all about pushing boundaries what's possible, even directly competing with gaming industry at a time: raycasters on low-end hw, voxel terrains, software raytracing etc... The art was there, just for the entertainment value - and frankly, quite kitschy art, which had it's charm. It's nice that more serious artists joined later on, but this it not exclusively artistic community in a slightest.
About opinions: as you can see, I'm not afraid of backlash, so go ahead. It's just more telling about you than anything else.
About opinions: as you can see, I'm not afraid of backlash, so go ahead. It's just more telling about you than anything else.
I think this was kind of an interesting experiment and the result is exactly what kb said. It's like a vantage point to a terrible future where art or nothing of significance exists at all, while also looking passable from a quick glance.
Quote:
"we can do cool things with AI" then fucking do cool things, this ain't it.
Please go ahead, do cooler things with AI by all means.
This was indeed just a low-effort attempt (which was the whole point really) to show what's possible with off-the-shelf tech used in a demoscene context. But as other said as well, it would be very interesting to push it to its limits.
Quote:
Please go ahead, do cooler things with AI by all means.
You are the one who has written a great deal of text on the BBS about how great AI will be for the scene, so you should put your money where your mouth is.
fizzer: it's more like I was one of the few who had a balanced opinion about it, as opposed to extremely polarized "are you with us or against us" type of view.
I think this is a very succesful experiment. At first I thought it was going to be a failure. Comments were coming in slow. But then it opened up. It gained way more attention than the “prod” by itself could ever deserve. Let me be clear, I think the prod itself, raken out of time and context of its release, is less than worthless. But the “performance art” experiment succeeded. Someone can code miracles, someone can draw masteripieces, someone can compose beautiful music. And someone can just enter a room of full people and fart loudly. Collect reactions. And that’s the art in this. You all are art. Your reactions are art. Tom’s trollish machine-like relentless halfwit blabber is art.
I can’t thumb this down, but I can certainly not thumb it up.
I can’t thumb this down, but I can certainly not thumb it up.
definitionally worthless, doesn't merit further thought
If you can't be bothered to make the prod, why should I bother to watch it?
About the low-effort thing, which seems to be now the main object of critique: sorry to break it down for you, but the view that art must involve visible labor, technical mastery, and so called "effort" is typical for working class aesthetic (Pierre Bourdieu). It makes a lot of sense why, if you think about it. So, no, in general low effort doesn't necessarily mean low value.
submit changes
if this prod is a fake, some info is false or the download link is broken,
do not post about it in the comments, it will get lost.
instead, click here !