Descriptions:
Product creator Peter Yang puts five lesser-known Google Labs AI tools through real-world tests in a focused 16-minute walkthrough covering Pomelo, Stitch, Project Genie, Flow, and the latest NotebookLM updates. Rather than a feature tour, Yang applies each tool to actual use cases and flags meaningful limitations—a practical counterweight to Google’s marketing framing.
Pomelo extracts a brand’s visual DNA and generates marketing campaigns; it excels for businesses with strong product photography (demonstrated on a local Asian bakery) but produces weak results for content brands lacking visual assets. Stitch, Google’s AI design tool, generates full design systems and app screens from a single prompt but skips the clarifying questions that competing tools like Claude Artifacts use, limiting quality for vague inputs. Project Genie creates keyboard-navigable interactive 3D worlds from text prompts in roughly 30 seconds—impressive as a tech demo but not a usable product. Flow, powered by Veo and Gemini in a unified canvas, earns the strongest marks as Google’s best current interface for image and video generation.
Yang closes with a broader strategic assessment: Google Labs produces capable underlying models but consistently ships standalone tools that lack integration with the broader Google ecosystem and trail competitors in polish and UX cohesion. For anyone tracking Google’s AI product strategy, this is a concise, opinionated field test of what’s worth trying today versus what remains firmly in demo territory.
📺 Source: Peter Yang · Published April 29, 2026
🏷️ Format: Review
