‘Google wants to combine search results and photos in personalized chatbot’

Google is reportedly working on Project Elmann, a way to build a personalized chatbot that tries to distill a person’s life story from search results, photos and videos and use them in responses.

Engineers are said to have presented the plan at an internal meeting, reports CNBC. Google says in a response that it is developing features with artificial intelligence and large language models and that it attaches great importance to privacy in new features. That is not a denial, although it is also not a confirmation of the existence of the project.

The intention is to be able to ask the personalized chatbot questions such as ‘when was the last time my brothers and sisters visited?’ and then get an accurate answer. The chatbot derives information through object recognition in photos and information that a person has given themselves.

The chatbot should also be able to deduce what a user would like to purchase or what possible travel plans are. Favorite sites and apps can also be deduced from screenshots, among other things, according to the slides of the project presentation. It is unknown if and when such a feature would come out. Google last week released Gemini, its series of language models for AI applications.