Google Gemini Generates Images from Photos
SkimNews Take
Integrating personal photo libraries directly into AI image generation transforms a user's past experiences into raw material for new creative output, effectively making individual life histories a direct input for the AI's imaginative capacity.
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google has enabled Gemini’s Personal Intelligence feature to connect with Google Photos for generating personalized images.
- Google Photos labels are used to identify people, friends, and family, which the Nano Banana 2 image model then visualizes.
- Personal Intelligence lets users input prompts such as “Design my dream house” or “Create a picture of my desert island essentials,” producing images that reflect their tastes and lifestyle.
- Google says it will not directly train its AI models on users’ private Google Photos libraries, though it will train on limited info like specific prompts and model responses.
- AI Plus/Pro/Ultra subscribers in the United States will receive the feature over the next few days, initially on Gemini on Chrome desktops, with broader rollout planned.
- Elijah Lawal, a Google spokesperson, described how the integration works and the privacy safeguards in place.
Why it matters: US AI Plus, Pro, and Ultra subscribers gain a new way to generate personalized images using their own photo data, while Google assures that the private library won’t be used to directly train its models, limiting privacy exposure but still using prompts and model responses for limited training.
