Over the past two years, AI video models have been competing on realism, resolution, and duration. But no matter how impressive the results look, we remain passive viewers: we press play, watch the clip, and it ends.
AlayaWorld
@alayastd is attempting something fundamentally different. Instead of generating a fixed video, it generates a world that continues to unfold as you move through it.
These three demos show the same journey toward a green village rendered in three distinct styles: photorealistic, oil painting, and line art. As the camera moves forward, the model continues generating the road, fences, trees, and distant village. This is not simply an existing video with different filters applied. The environment is generated continuously along the camera trajectory, allowing the scene to develop as the user explores it.
AlayaWorld streams video at 720p and 24 FPS while supporting camera movement and viewpoint control. The real breakthrough is not just image quality. Once generation becomes fast enough to respond within an interactive loop, the user is no longer merely watching a video. They become a participant inside the generated world.
The world can also respond to new instructions. During generation, users can introduce prompts that trigger spells, summon characters, create explosions, or transform the environment. Most video models follow an initial prompt and produce a predetermined clip. AlayaWorld can respond to changing intent while the world is still running, allowing subsequent events to evolve according to the user’s commands.
Generating an attractive frame is relatively easy. Maintaining a coherent world over time is much harder. As a video model repeatedly predicts the next frame, small errors can accumulate until roads, buildings, and objects begin to distort or disappear. AlayaWorld combines spatial memory with compressed historical context, helping the model remember both where things are and what has already happened. This enables stable generation lasting more than one minute while improving consistency when the camera leaves an area and later returns.
This may be the next step for AI video: not simply generating a longer movie, but generating a world that can be explored, changed, and interacted with.
AlayaWorld is developed by Alaya Lab. The team is progressively releasing its inference code, training code, and datasets, with an online experience expected to launch near the end of the month.
Project page:
显示更多
Seedance 2.0 is pretty wild
100% AI
What if Jurassic Park was real?
PROMPT:
Format: Found-footage vacation video, 15 seconds, seven shots with hard cuts — a tourist's camera roll from one full day, morning to dusk.
Image reference: Use the attached gate photo as the exact visual reference for Shot 1 — match its architecture, proportions, torch placement, lighting, and signage precisely.
Main subject: Young woman, blond hair in a loose low bun shedding strands, oversized faded olive t-shirt, denim shorts, white canvas sneakers, disposable-camera strap on her wrist. Realistic skin texture, minimal makeup, wide-eyed tourist joy. Maintain perfect visual consistency — identity, hair, clothing, proportions — across every shot and location.
Location: A vast island dinosaur preserve. Jungle mountains, open savanna paddocks, wooden observation structures, tall electrified perimeter fences, dirt tour roads. Dinosaurs are photorealistic animals — muted natural hides, birdlike motion, visible breathing weight. Filmed like wildlife, never like monsters.
Camera style: Mid-range smartphone, handheld by her friend. Heavy shake, autofocus hunting, motion blur on pans, compression grain, exposure pumping. Each cut is abrupt and mid-motion, like real unedited camera-roll clips.
00:00–00:03 — THE GATES (morning, filmed from inside a moving open-top tour jeep). The camera films forward over the seats as the jeep approaches the colossal timber gates from the reference image — flared stone-and-wood towers, flaming torches, jungle pressing in on both sides, the road still wet from morning rain. The doors groan open. She spins around from the front seat into frame, gripping the roll bar, and half-shouts over the engine: "Oh my god, this is wild! I can't believe we're actually here!" — as the jeep passes under the archway into shadow and out again.
00:03–00:05 — SAVANNA (late morning, wide shot from a raised wooden platform). She stands small at the rail against an open plain where a herd of long-necked giants wades through heat haze, one rearing slowly against the treeline. The camera zooms in shakily, overshoots, corrects.
00:05–00:07 — RIVER CROSSING (midday, low angle through the jeep's open side). Past her shoulder as the jeep fords a shallow river: a crested duck-billed dinosaur drinks twenty meters away, lifting its head to watch them pass, water dripping from its jaw. She raises her disposable camera; the phone catches its tiny click.
00:07–00:09 — JUNGLE FENCE (afternoon, handheld walking shot). Dim under the canopy beside a tall humming fence. She peers through the wire — and a small crested predator peers back from the ferns, head tilting in quick birdlike jerks. She backs up a slow half-step, whispering "okay... hi." Autofocus argues between wire and eye.
00:09–00:11 — THE HERD RUNS (late afternoon, chaotic pan from the jeep at speed). A flock of small striped dinosaurs floods across and around the vehicle. The camera whips left-right, blurred bodies streaking past, her whooping off-mic, dust on the lens.
00:11–00:13 — QUIET GIANT (golden hour, static shot at a wooden overlook). The frame finally settles: silhouetted against low sun, she stands at a rail as an immense horned dinosaur grazes just below, close enough to hear grass tearing. She isn't laughing anymore — just watching.
00:13–00:15 — DUSK ROAR (blue hour, half-framed and accidental). Filmed as the phone was being lowered, frame tilted: a colossal, deep bellow rolls in from the darkening jungle far off. Everyone freezes and turns toward it. Cut to black mid-turn.
Audio: Ambient only, shifting per location — jeep engine and gate groan, her excited line over the motor, savanna wind, river wading, fence hum, thundering small feet, grass tearing, and the single distant bellow that swallows everything. No music. No narration.
Style & quality boosters: Raw consumer camera-roll aesthetic, abrupt mid-motion cuts, heavy handheld instability, natural motion blur, golden-hour lens flare, coherent animal mass and physics — ground tremor, water displacement, dust. No text overlays, no watermarks.
Goal: One unforgettable day inside the world's most famous dinosaur preserve, told through seven imperfect tourist clips — opening on the gates everyone recognizes, and closing on the sound that reminds them what lives beyond the fences.
显示更多