PHGoogle just unveiled Gemini Omni, an AI model that can generate anything from any input — starting with video. You can type a description, upload an image, or even give audio, and it produces a realistic video output.
It’s physics‑aware (gravity, energy, fluid dynamics), so scenes look natural. Plus, every video is digitally watermarked for authenticity.
Meanwhile, Stanford reports AI models now solve PhD‑level science problems but still struggle with simple tasks like reading analog clocks.
AI is evolving fast — from coding super‑speed to full multimodal creativity. Imagine describing “a typhoon hitting Manila” and instantly seeing a simulation video.
It’s physics‑aware (gravity, energy, fluid dynamics), so scenes look natural. Plus, every video is digitally watermarked for authenticity.
Meanwhile, Stanford reports AI models now solve PhD‑level science problems but still struggle with simple tasks like reading analog clocks.
AI is evolving fast — from coding super‑speed to full multimodal creativity. Imagine describing “a typhoon hitting Manila” and instantly seeing a simulation video.