AI 图像生成正在变得更好。

AI image generation is getting better. But this feels like a real leap forward. I tried Uni-1 by @LumaLabsAI, and the output wasn’t just “good”… it felt intentional. Not just objects placed together, but a full scene with mood, lighting, and coherence. Prompt I used: A group of cute cartoon animals (cat, dog, panda, fox) intensely playing video games in a cozy neon-lit room, glowing screens reflecting on their faces, snacks scattered around, playful chaos, vibrant colors, animated movie style, ultra-detailed What stood out: • Lighting behaves realistically across characters • Composition feels cinematic, not random • Details stay consistent even in complex scenes • Chaos is controlled, nothing feels broken This is where things are heading: From generating images → to generating experiences inside a frame Quick try: Go to https://t.co/9HWXlMQoFy Create a board Select Image → Uni-1 Paste prompt → generate This is what “prompt → world” should feel like. 🚀

AI 图像生成正在变得更好。 但这一次感觉像是真正的飞跃。 我试用了 @LumaLabsAI 的 Uni-1，输出不仅仅是“好”……它让人觉得是经过精心设计的。 不只是物体被简单地放在一起， 而是一个完整的场景，有氛围、有光线、有连贯性。 我使用的提示词： A group of cute cartoon animals (cat, dog, panda, fox) intensely playing video games in a cozy neon-lit room, glowing screens reflecting on their faces, snacks scattered around, playful chaos, vibrant colors, animated movie style, ultra-detailed 最让我印象深刻的地方： • 光线在不同角色之间表现得非常真实 • 画面构图有电影感，而不是随机的 • 即使在复杂场景中，细节也保持一致 • 混乱是受控的，没有任何崩坏的感觉 这正是未来的方向： 从生成图像 → 到在一个画面内生成体验 快速试用： 前往 https://t.co/9HWXlMQoFy 创建一个画板 选择 Image → Uni-1 粘贴提示词 → 生成 这就是“提示词 → 世界”应该有的感觉。🚀