Gemini is Google’s next-generation large language model (LLM). According to a new report, Google may use Gemini to power a new AI feature that can “answer previously impossible questions” about your life using phone data such as photos and searches.
Google Working on a New AI Project Called Project Ellmann
In a major Android updates today as reported by CNBC, Google is developing an AI experience called “Project Ellmann.” Project Ellmann, named after biographer Richard David Ellmann, would use LLMs to gain awareness of what’s going on in a photograph. According to reports, Google wants Project Ellmann to become Your Life Story Teller. It’s unclear whether the chatbot will be integrated into Google Photos or any of the company’s other products.
The feature appears to work by collecting search result data and detecting patterns in your photos. Project Ellmann would then be able to describe a user’s photos in greater depth by using biographies, previous moments, and subsequent photos rather than “just pixels with labels and metadata,” according to the outlet.
Presentation
The presentation appears to include a few examples, such as someone who recently attended a school reunion. Project Ellmann is said to be able to deduce that the user graduated ten years ago and that the photo includes faces they haven’t seen in a decade, implying that the photo is a reunion photo.
Conclusion
Google believed to be also demonstrated “Ellmann Chat,” with the description, “Imagine opening ChatGPT but it already knows everything about your life.” “How would you put it?” In response to the question “Do I have a pet?” the chatbot was able to answer yes while also providing additional information about what the dog was wearing, the dog’s name, and the family members it spends the most time with.