Try out SAM 3 here: https://aidemos.meta.com/segment-anything I’ve spent the last few weeks testing Meta’s new SAM 3 (Segment Anything Model). I have been using it to add video effects, create presentations, and build 3D scenes from images. In this video, I break down the five main tools Meta just dropped in the SAM 3 Playground. 1. Video Tracking: Prompts to select and add effects on moving objects. 2. Auto-Counting: Identifying hundreds of items in a single click. 3. 3D Scenes: Creating digital assets from real-world products. 4. 3D Bodies: Turning human movement into 3D skeletons. 5. Audio Isolation: Cleaning up messy background noise instantly. 0:00 - Intro to SAM 3 0:46 - Video Cutouts: Using text prompts to track and add effects (Glitch, Trail, Clone) 2:25 - Image Cutouts: Automated object counting (Umbrellas & Scientific Cells) 3:52 - 3D Scenes: Turning 2D product photos into 3D assets (.PLY / .GLB) 4:42 - 3D Bodies: Human pose estimation for fitness and gaming 5:16 - Isolated Sounds: Removing background noise and adding "Studio" & "Lo-fi" effects 6:40 - Coming soon to Instagram & Vibes

How I Build Real Apps Without Writing Code | Lovable Tutorial
268 views

The Pope Wants to "Disarm" AI
4.9K views

Grads boo AI, Reese Witherspoon gets dunked + Karpathy joins Anthropic | TWiAI E14
6.5K views

How to Use GPT Image 2: 4 Real Workflows That Actually Work
622 views

How to Get Consistent Brand Designs From Any AI Agent
397 views

Hermes Agent Just Made OpenClaw Look Old
1.4K views