The Google team proposed an idea to use AI technology to process users** and search engine query information to comprehensively analyze what the user's life is like, and the project is called "Project Ellmann".
As originally envisioned, Project Ellmann would use LLMs (such as Gemini) to extract information, find specific patterns from the user's **, and then build a chatbot that would answer questions accurately. Google claims that Ellmann will be a "personal life storyteller."
Sounds interesting, but does Google want to embed Ellmann functionality into Google Photos, or does it add it to other products? It is not yet known. Google Photos has more than 1 billion users and stores 4 trillion ** or **.
Google is currently using AI technology to optimize its product line, and Ellmann is just one of the ways to do it. Google launched Gemini a few days ago, and on some occasions Gemini has even surpassed OpenAI GPT-4.
Google intends to license Gemini to Google Cloud users, who can use Gemini to develop the features they want. The most prominent advantage of Gemini is that it is "multimodal", which means that it can understand many types of information such as text, audio, etc.
At a recent internal meeting, an executive from Google's Photos division presented Project Ellman. Internal documentation shows that the Google team believes that a large language model can be used to provide a "bird's-eye" view of a user's real-world life story.
Through biographies, previous records, **, and other materials, Ellman is able to understand the situation deeply. For example, by analyzing a series of fragments, Ellman can know whether the subject of various messages is college life.
In the presentation slides, one reads, "We can't answer difficult questions and tell a good story without a bird's-eye view of our personal lives." ”
Another says, "We'll look at your ** to see what its label and location are, and then determine if they belong to a meaningful moment." When we fully understand your life, the personal story becomes clearer. ”
The Google team also talked about Ellmann Chat. What is Ellmann Chat? The team described: "Imagine what questions you would ask it when you opened it and it already knew everything about your personal life. ”
For example, a user asks Ellmann Chat a question: "Do I have a dog?" The AI replied that the owner had a dog, and also said the name of the dog, and said that there are two people in the family who like to be with the dog the most, and what is the name of the family.
Another problem is that the owner wants to move, so the AI can provide some small towns to choose from that that are similar to the existing living environment. Ellmann can answer these questions. Not only that, but Ellmann also knows what the user's diet Xi are like.
For example, what users want to buy, what interests they have, what work and travel plans they have, Ellmann can analyze them and find the answers. What users like to visit and what apps they use, Ellmann is even clearer.
A Google spokesperson revealed: "Google Photos has always wanted to use AI technology to help users search for **and**, and with LLM, we can provide a more practical experience." Ellmann is still in the early stages of internal exploration, and it will take time to determine whether it will actually be launched. Google also has to consider the privacy and security of its users to make sure that Ellmann is really useful to users. ”
Project Ellmann is just one of many tech companies that want to use new technology to create more personalized personal memories for their users.
Whether it's Google Photos or Apple Photos, they're all analyzing**, finding patterns from them, and then making photo albums. Google Photos can automatically group similar** together into an album for users to find. Apple said in June that after the software is upgraded, it can identify the person, dog or cat in the photo, and the user only needs to say the person's name, and the software can help you find his**.
It sounds wonderful, but Google and Apple's technology is still imperfect and sometimes makes mistakes. In 2015, Apple's Google AI made a joke about identifying black people as chimpanzees, and even today, mistakes are inevitable. (Knife).