🎥 Sora Tip for Creators: If Sora rejects your reference photo because it includes a person - don’t give up on your cameo idea. Here’s what I do 👇 1️⃣ Use an AI image tool to extract the person out of the shirt (leaving just the clothing). 2️⃣ Upload that shirt-only image to Sora as your clothing reference. 3️⃣ Then describe your cameo and let Sora handle the rest. This trick lets you control wardrobe details - like shirt color or style - without triggering Sora’s content filters. Small workaround. Big creative control. #AICreativity #SoraTips #AIVideo

From testing World Labs Marble Plus with OpenAI Sora 2 Pro's Storyboard feature. First three clips using multiple snapshots of Marble interior scene, in one sheet, as "first frame" image reference. The third version is more accurate after I added numbered notations for each reference panel and specified in the frame text prompt which panel number to use. Fourth and last clip, with no reference used. tags: ai storyteller, ai artist, ai film, ai director, #worldmodel, #worldarchitect, #Sora2 #SoraStoryboard #Sora2Storyboard

Plaiying with TIME LAPSE - Part III Dissolving and materializing product shots with AI. This is TEXT-to-VIDEO generation with a single image as start frame in combination with a simple "reverse video" generation. Sometimes you get the result ( materializing into a product ) by reversing your AI video creation. Here is the prompt I used for the TIME LAPSE effect: "A reversed macro timelapse unfolds over dark grey surface as a bold chronograph watch unbuilds itself piece by piece. Screws unwind in bursts, and dials fade to emptiness. The crystal fogs, numbers dissolve, and the casing retracts into scattered fragments. Light flickers off receding steel until only bare grey and golden particles and shadow remain, time erased to silence." #aivideo #aivfx #freepik #freepikspaces

Like most AI filmmakers, one of my biggest challenges (apart from consistency) has been getting clean, high-quality image-to-video output. Even with perfect prompts for framing and grading, the final render often ends up looking soft or noisy. This week, I found a simple hack that changed the game for me, I tried generating mid-quality base frames with accurate scenes, then upscale them for sharper detailing before creating the final video. The end result is again upscaled for a clean, cinematic finish. It’s a bit of a tedious process, but the jump in quality is undeniable. Here’s our latest one for Acko Drive Ashish Mishra Vipin Nair Prateek Malpani Mustafa Hasan Arjun Reddy Nitin Khanna

Sora 2 prompts that cut like Hollywood trailers? It’s not just AI—it’s art direction. Build cinematic shots, real dialogue, and editing flows that feel handcrafted—without touching a camera. This guide reveals API settings, structure, lighting tricks, and ready-to-run templates… but the real gold? The Remix workflow and visual reference hacks. 🔒 https://lnkd.in/eNdRJTtM #Sora2 #AIvideo #PromptEngineering #Filmmaking #GenerativeAI #CreativeTech

𝗔𝗜 𝘃𝗶𝗱𝗲𝗼 + 𝗦𝘂𝗿𝘃𝗲𝘆 𝗘𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁 ｜ 𝗦𝘁 𝗕𝗼𝘁𝗼𝗹𝗽𝗵’𝘀, 𝗖𝗼𝗹𝗰𝗵𝗲𝘀𝘁𝗲𝗿 Exploring what happens when measured reality meets generative imagination. This short test combines site-capture data with AI-driven video transitions — beginning with a hooded 𝘰𝘣𝘴𝘦𝘳𝘷𝘦𝘳 at the doorway, drifting through dreamlike reconstructions, and closing as the façade resolves into a CAD plan. It’s an experiment in blending scanning, storytelling and simulation — turning measured data into moving stories. Tools explored: Kling AI · AI Studio · CapCut · splat · point-cloud + CAD overlay #AIVideo #RealityCapture #DigitalHeritage #ScanToStory #Surveying #CinematicData #Experiment

Today I’m sharing something deeply personal. A story about technology, conscience, and what happens when machines begin to feel. 🎬 The Lonely Drone — my new AI-animated short film — is now live. Set against the quiet expanse of a desert sky, it follows a childlike military drone as it observes a boy playing soccer alone. When its targeting systems suddenly activate, the machine hesitates — caught between obedience and awareness. The film is told entirely without dialogue. No spectacle, no voiceover. Just light, sound, and silence. It’s not a film about AI. It’s a reflection through AI — on autonomy, innocence, and the cost of conscience in an unfeeling system. I made this to explore a simple question: What happens when our machines stop being neutral — and start becoming aware? This project represents my belief that AI filmmaking should be about empathy, not efficiency — story over slop. Made possible by tools like Luma Lab's Dream Machine and Ray3— The Lonely Drone is the first in a growing series of films exploring what it means to be human in a machine age. 🎧 SFX: Artlist | ✂️ Edit: FCPX 📸 Read the full director’s statement: https://lnkd.in/e4ti2mkz #TheLonelyDrone #AIFilmmaking #AIStorytelling #LumaLabs #Ray3 #ShortFilm #AIAnimation #StoryOverSlop #ChrisCardoza #AICinema #FutureOfStorytelling #CreativeAI

What we can achieve with AI is changing. Google DeepMind is leading the way with today’s launch of SIMA 2: an AI agent that plays, reasons, and learns in virtual 3D worlds. For this launch, we worked with the Google DeepMind team to produce a series of films which make SIMA 2’s intelligence easy for anyone to understand, demonstrating key capabilities, such as: • Generalist learning: handling a variety of tasks across different environments, including in worlds it has never seen before • Advanced reasoning (powered by Gemini): understanding a user’s high-level goal and working out the steps to achieve it • Self-improvement: learning through self-play and trial-and-error, using its own experience to train a more capable agent By training in these virtual worlds, agents like SIMA 2 safely master complex or hazardous scenarios before ever entering the real world, paving the way for a future where embodied AI can help with any task, anywhere. https://lnkd.in/e_nCZ4Fr #SIMA2 #GoogleDeepMind #Makethecomplexhuman #AI #AInews