Alisa Qian's banner

Alisa Qian

@alisaqqt • 2,010 subscribers

Product Marketing Lead | Building AI Infra @atlas_cloud_ai | ex @novita_labs | Bay Area ⚡️ Your source for AI insider intel

Shorts

Seedance 2.0 is amazing. Honestly, just let AI handle the next seasons of Doctor Who and Stranger Things. Don't like the plot? Just generate your own. 😂 The future of streaming is "Prompt-to-Watch". Follow me and get notified once Seedance 2.0 dropped.

Seedance 2.0 is amazing. Honestly, just let AI handle the next seasons of Doctor Who and Stranger Things. Don't like the plot? Just generate your own. 😂 The future of streaming is "Prompt-to-Watch". Follow me and get notified once Seedance 2.0 dropped.

37,033 Aufrufe

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

I made a Vox-style explainer about Tang Dynasty. Just one prompt, 100% AI / end to end. keyframes (Seedream 5.0 Pro) → animation (Omni Flash) → voice (xAI TTS) → music (MiniMax Music 2.6) → assembled locally on Claude Code. The whole pipeline is on Atlas Cloud. One topic in, finished film out.

I made a Vox-style explainer about Tang Dynasty. Just one prompt, 100% AI / end to end. keyframes (Seedream 5.0 Pro) → animation (Omni Flash) → voice (xAI TTS) → music (MiniMax Music 2.6) → assembled locally on Claude Code. The whole pipeline is on Atlas Cloud. One topic in, finished film out.

110,567 Aufrufe • vor 20 Tagen

Seedance 2.5's reference capability is insane. Open source is dead.

Seedance 2.5's reference capability is insane. Open source is dead.

43,213 Aufrufe • vor 1 Monat

I typed one sentence — "introduce Mexican food, 60s, 16:9" — and Claude Code gave me this back. Now I'm open-sourcing this skill that turns any topic into a finished Vox-style paper-collage video. Install it, ask for a video, get an mp4.

I typed one sentence — "introduce Mexican food, 60s, 16:9" — and Claude Code gave me this back. Now I'm open-sourcing this skill that turns any topic into a finished Vox-style paper-collage video. Install it, ask for a video, get an mp4.

20,412 Aufrufe • vor 18 Tagen

Seedance 2.5’s most underrated industry upgrade is its 3D white model input rendering — worth a close look for anyone in sci-fi film & VFX. It fixes real pre-production pain points, with clear efficiency wins: - Faster previs iteration: Get atmospheric previews straight from locked layout, no waiting on lighting & texture teams. Cheaper creative revisions. - Lower costs for pitches and concept validation. No wasting senior team hours on bid demos. - Earlier reference for VFX & art teams, extending cross-department parallel work. - Fast turnarounds for secondary shots (distant setups, alternates), saving render time and manpower. Official demo: ~100k-face spacecraft animation + style references → fully rendered footage. Structure and proportions hold steady in dynamic shots, and your original layout stays intact. One clear limit: it is still a generative previs tool, and cannot yet replace industrial physically accurate rendering and standard asset pipelines. It won’t take core lighting & rendering jobs, but it will reshape how previs gets done.

Seedance 2.5’s most underrated industry upgrade is its 3D white model input rendering — worth a close look for anyone in sci-fi film & VFX. It fixes real pre-production pain points, with clear efficiency wins: - Faster previs iteration: Get atmospheric previews straight from locked layout, no waiting on lighting & texture teams. Cheaper creative revisions. - Lower costs for pitches and concept validation. No wasting senior team hours on bid demos. - Earlier reference for VFX & art teams, extending cross-department parallel work. - Fast turnarounds for secondary shots (distant setups, alternates), saving render time and manpower. Official demo: ~100k-face spacecraft animation + style references → fully rendered footage. Structure and proportions hold steady in dynamic shots, and your original layout stays intact. One clear limit: it is still a generative previs tool, and cannot yet replace industrial physically accurate rendering and standard asset pipelines. It won’t take core lighting & rendering jobs, but it will reshape how previs gets done.

17,565 Aufrufe • vor 1 Monat

I deleted half of my AI-video prompt — and the result got better. 😶 Here's the before/after that changed how I prompt, and the skill I open-sourced from it. 🧵 We've been prompting video models like a film shot list: 24mm, f/1.4, "volumetric fluid simulation," frame-by-frame timing. But a lot of the time you don't need to — the smarter the model gets, the simpler the prompt can be: tell the story, the texture of the air, the emotion, and let it pick the shots, light, and rhythm. ByteDance shipped this idea with Seedance 2.0 and called it "Vibe Creating" — I open-sourced the skill and the philosophy behind it. Same scene, two prompts 👇 🔧 Regular: "85mm f1.4 macro, 120fps, dolly 0.6x, freeze the sweat at 1/250…" ✨ Vibe: "Late-night street stall. The cook flicks the wok; a ball of orange flame lights up his sweating face. Noodles fly. He plates them and wipes his brow." Same model. One of them feels alive. The skill is built for story-driven video — concept shorts, micro-narratives, emotional or atmosphere pieces, anything where you'd rather describe the moment than dictate the camera. Feed it your idea (or your over-stuffed prompt) and it hands back the version the model actually shoots better. The one place it won't go: UI demos, step-by-step tutorials, or exact dialogue sync — there it'll tell you it's the wrong tool rather than flatten a prompt that needed to stay precise. Try it on your next prompt — repo 👇

I deleted half of my AI-video prompt — and the result got better. 😶 Here's the before/after that changed how I prompt, and the skill I open-sourced from it. 🧵 We've been prompting video models like a film shot list: 24mm, f/1.4, "volumetric fluid simulation," frame-by-frame timing. But a lot of the time you don't need to — the smarter the model gets, the simpler the prompt can be: tell the story, the texture of the air, the emotion, and let it pick the shots, light, and rhythm. ByteDance shipped this idea with Seedance 2.0 and called it "Vibe Creating" — I open-sourced the skill and the philosophy behind it. Same scene, two prompts 👇 🔧 Regular: "85mm f1.4 macro, 120fps, dolly 0.6x, freeze the sweat at 1/250…" ✨ Vibe: "Late-night street stall. The cook flicks the wok; a ball of orange flame lights up his sweating face. Noodles fly. He plates them and wipes his brow." Same model. One of them feels alive. The skill is built for story-driven video — concept shorts, micro-narratives, emotional or atmosphere pieces, anything where you'd rather describe the moment than dictate the camera. Feed it your idea (or your over-stuffed prompt) and it hands back the version the model actually shoots better. The one place it won't go: UI demos, step-by-step tutorials, or exact dialogue sync — there it'll tell you it's the wrong tool rather than flatten a prompt that needed to stay precise. Try it on your next prompt — repo 👇

11,501 Aufrufe • vor 1 Monat

Keine weiteren Inhalte verfügbar