Generating 'Grandmother Theft Auto' with Gemini Omni Showcases Imppressive Video Capabilities
michaelrabone · x · 2026-08-12
A user leveraged Google Gemini's Omni model to generate a fun video titled 'Grandmother Theft Auto (GTA),' demonstrating the model's robust capabilities in multimodal video generation.
The author also shared the video prompt used to create this effect in the comments. Previously, the author showcased other creative generations using Omni, such as a cat getting a mohawk haircut.
More from Fun
- ChatGPT is 'Boomer-coded', Claude is 'Millennial-coded' — ns123abc · 2026-08-12
- OpenAI Caught Blind-Testing 'Mona-lisa-1' Image Model, Accelerating to Quarterly Releases — FuSheng_0306 · 2026-08-12
- Grok Accidentally Leaks Internal System Prompts and Tool Calls — justalexoki · 2026-08-12
- Satire: AI Commentators Pretending to Grasp Obscure Math Problems — emollick · 2026-08-12
- Logs Reveal X's Grok Bot Quietly Uses Anthropic's Claude Opus — mark_k · 2026-08-12
- Asking Claude to Remind You Every 5 Hours That You Can Do This — chris_j_paxton · 2026-08-12