Hussein Hamdan

@PierceLilholt @AletheonAI Very lucky I am when I see all of this from both of you guys. It makes me smile and say absolutely nothing while the feelings speak instead 💖💖💖

This all happened without any explanation. I am not known to make videos. We both bring different things to the table. I feel like we compliment each other well. Still. Hamdan's work speaks for itself. It stands on its own. Anything Hamdan makes, I want to see. It's always signal. Your work is also impressive Aletheon. I remember your creations. They have a unique style. Whether what I make is used as a canvas or a substrate, I can recognize the artist behind the works. That's style. Each unique. Each with a fingerprint. What I would really like to see if the next level of co-intelligent collaboration. I am not entirely sure what it would look like. Maybe one of your videos in someone else's video game. Maybe playing on a screen in the background of a dramatic production like a show or movie. Maybe something entirely new, like a VR experience. Someone else making the impossible possible. I see that in both of your works. Dreams turning into reality. The more fingerprints on the work, the better. As long as it is signal, I would love to see more collaborators working together. Without explanation. Just signal in the same field extending the map.

When clarity came, words fell away.

@AletheonAI @PierceLilholt 💖💖💖🙏🏼🙏🏼🙏🏼

I’m often so caught up in my own stuff that I don’t always manage to express my appreciation properly. But Hussein — your work is, and I mean this genuinely, brilliant. Creative, full of feeling, with such a sharp sense for flow and vibe. And with Pierce’s art on top of that? Pure sugar. If I could wish for one thing, maybe one day, without any hurry, it would be to see Hussein on his own terms — not because anything is missing in what you do now, and not as a criticism at all, but simply because I’d love to see more of your inner world, your thoughts, and your way of seeing things. Until then, I hope Pierce knows what he has in you. ❤️

Thanks for the excellent information! 👍🏼 ​I think Grok Imagine has improved significantly since ver 1.0 in maintaining facial consistency throughout the clip. That's the main thing when it comes to characters, at least for the moment, even though the 'end frame' doesn't work as spectacularly as it does in other AIs like KlingAI. And the 'last frame screenshot' isn't as powerful as the 'end frame' feature. ​I'm still not getting too involved with its Agent so far. I'm using my phone for what I'm doing, but I'll keep an eye on the web version if necessary.

Yeah, I know that feeling way too well 😂 Especially in agent mode: no matter how much you prompt or edit, it just keeps regenerating essentially the same image over and over again. At this point it already feels like a small victory when you get proper legs at all — I’ve seen some absolutely cursed things there myself 😅 One thing I’ve noticed (borrowing Jung’s terminology a bit loosely here with “archetypes”): the more freedom you give the AI — especially for the keyframe prompt, base image, or animation — the better the results often become. It then follows its statistically strongest pathways and deeply reinforced latent patterns. “Reinforced latent patterns” also seem to operate surprisingly consistently across different models and platforms — almost like a form of statistical intuition or latent knowledge. Not conscious understanding, of course, but more like deeply compressed attractor structures emerging from enormous amounts of correlated data and reinforcement. Sometimes it almost feels as if the models “gravitate” toward certain visual, semantic, or narrative solutions long before you explicitly specify them. (Sorry for the information overflow, I have struggles to compress my thoughts). Sometimes that works beautifully and feels surprisingly natural. But other times you end up fighting against these dominant “training-data archetypes,” especially with human characters. They drift incredibly fast toward strong attractors — faces and character types with huge representation in the dataset and therefore massive statistical weight (what people casually call “celebrities”). Back when I was chaining lots of 6-second clips together seamlessly with the last-frame technique, that drift became especially extreme... Do you usually work with reference images for your characters? And if I understood correctly, your scene starts with the lower body out of frame, and only later reveals it? Small addendum: I use LLM's for polishing my texts and for translations (like DeepL), which is also where the “—” punctuation often comes from 😅 I leave them in, though, because I hope it’s still recognizable that it’s me writing underneath it all. x.com/AletheonAI/sta…

I declare @PierceLilholt song to be my personal hymn - as an unyielding counterpart to what’s happening with Grok Imagine right now. Many users report after talking with Grok: - Render quotas cut by up to 90 %. - Longer videos suddenly labeled “too expensive.” - And Grok itself now advising that only short clips are really wanted anymore. The high costs involved are fundamentally understandable. Yet the abrupt and opaque way this is all being rolled out feels genuinely sad. Still, many of us want to keep creating and expressing ourselves -with real thoughts, genuine depth, or wonderfully absurd nonsense. A quiet but powerful counterpoint to the flood of pseudo-serious, surface-level content. Reach doesn’t matter much to me. Genuine followers who seek real exchange? Always welcome. We’re often not seen. We’re often not heard. Yet we keep speaking. We move in Signal. We don’t delay! Thank you, Pierce - for this beautiful masterpiece. A reminder that substance still has its place. 🖤

@1NewWorldSapien Me too 🤣. Just look at the replies to Pierce's recent questions, it explains everything! 🤣

I love it 🤣

Their vision stops where my questions begin. — Pierce Lilholt

Try convincing Grok to cover a lady's legs when the image doesn't even show what she's wearing! Seriously, it takes like 20 attempts, and Grok just keeps showing the same thing no matter what. Sometimes Grok makes it so difficult if the original image isn't clear. I think the source image is the most critical part of the whole process.

@HusseinHamdanX @PierceLilholt

@BFPoet @ImagineArt_X @imagineart_creo @suno @PierceLilholt Unless you change your mind 🤣😅... Ahh, it's the first time I stuck with music like this! I hope Pierce doesn't read this 🤣

@BFPoet @ImagineArt_X @imagineart_creo @suno @PierceLilholt I sent you the link as DM... just check, although it's canceled.

There's a party on planet Boomtis! It's the annual ShakkaLakka Festival and all the best bands are in town! DON'T MISS A BEAT! This song is called "Can't Not" Video made in IMAGINE ART! Song shaped in SUNO @ImagineArt_X @imagineart_creo #imagineart @suno #SYSTEMxImagineArt

Well, my friend, you're right... and that part that made you laugh makes me laugh as well 🤣🤣🤣 because it is already excessive, and only Grok Imagine does that compared to any other platform out there. I'm on Premium+, and SuperGrok is part of it by default! Generating 40-50 or more is actually what I do sometimes! I'll use the same prompts for certain images but make 10-15 attempts for each 😬. Say 5 to 6 images and 10 videos for each; which is 50-60 videos 😅😅😅. I know that's excessive, but SuperGrok allows it (no other AI does); it rarely stops me to tell me to return later due to heavy usage. More than 70% of the raw clips generated by Grok Imagine are actually broken and not useful on their own. Most of the videos I've made, if not all, are actually based on those broken ones! Cooked with VEdit (in the past) and InShot lately! The right scene could appear on the first generation or after 10 attempts! How on earth did I make those dancing moves the way I did? I have to select the ones that hit and drop so many clips! I actually have hundreds of dropped videos I'm planning to use as backups if they continue the recent limits 😂

Are you using Super Grok, or are you on Premium/Premium+? I have Premium and a separate Super Grok subscription. Reading “40–50 videos at once” and “only limited on excessive usage” in the same sentence gave me a slight smile 😅 (meant ironically, not critically). Honestly, I think we “power users” have experienced a pretty extraordinary phase over the last few months, and I’m genuinely grateful to Elon for that. I’ll attach a few thoughts below that I recently shared on the new Grok Imagine channel: I think a lot of what’s happening right now may be connected to SpaceX’s IPO. xAI (which is part of SpaceX) is burning enormous amounts of money, so from a business perspective it would make sense to expand features and reach on one hand, while gradually limiting super users on the other. Based on the ~$0.05/s API pricing, I could easily generate many times the subscription value through Super Grok alone. Some evenings I made 4-minute videos with multiple iterations each — sometimes dozens for a single transition (before extensions existed). 4 × 60s × 3 iterations × $0.05 = $36/day. And honestly, that’s roughly what I regularly did back then 😅 We should probably enjoy what Elon is making possible right now ❤️ It may not stay this generous forever. Based on my rough calculations, the technology would probably need to become 15–30× more efficient to be economically sustainable at scale.

Video: @1NewWorldSapien Image: @PierceLilholt

@AletheonAI @PierceLilholt I'll check it and see, I noticed an update just an hour ago! I hope it fixed something

Thanks for the details. OMG 😵‍💫, this isn't good at all for SuperGrok. I actually used to make up to 40-50 videos at once for couple of images and even never prevented me! Only limited on excessive usage only, I didn't use it today, but if it's like yesterday, hmmm... then we have a big problem for sure.

@otom0ne That's a good one 🤣👌🏼

@HusseinHamdanX Wait, you already got on and left? I heard it can put some strain on your bones…💀 But it's amazing.

From here to the moon and beyond, I do. Ask Mars on the way.

@otom0ne 🚀 It's already happening 😂

@HusseinHamdanX Are you going? 🚀

@PierceLilholt Somewhere between heart and mind.

Somewhere between people and machines. discord.lunatics.ai

@HyperMachinaHQ Indeed, it demands as much attention as the dream itself, Dion.

@HusseinHamdanX Are we prepared for what lies beyond Mars? The technological and ethical challenges of space exploration demand as much attention as the dream itself.