This Gemini calendar hack shows why Google is uniquely placed to win in AI (www.businessinsider.com)

A Business Insider writer tried Google’s Gemini app to turn screenshots of a bocce league schedule into real Google Calendar events. They uploaded phone screenshots, told Gemini “add these events to my Google calendar,” and the app linked to their calendar, parsed dates/times/locations, and created accurate events in about 25 seconds. The feature felt “magical” because it eliminated manual copy-paste and repeated invites, demonstrating a smooth multimodal workflow from image input and natural-language instruction to concrete calendar actions. For the AI/ML community this is a compact proof point of what multimodal assistants can do: OCR and information extraction from images, intent understanding, and secure API integration to drive side-effectful actions (creating events). It also highlights a strategic advantage for Google—owning a widely used calendar and identity stack lets it productize these capabilities more readily than rivals (OpenAI doesn’t offer a calendar), creating a practical moat around everyday automations. The example underscores both opportunity (automating fiddly tasks through LLMs) and considerations—auth, permissioning, and privacy when models act on personal data—while hinting at how tightly integrated services could accelerate adoption of assistant-style AI features.