Google is developing an AI assistant that will be able to answer ‘impossible’ queries about you

Google is developing an AI assistant that will be able to answer ‘impossible’ queries about you.Google is apparently working on an AI assistant that can examine personal images, files, and Search results in order to tell “your life story.”

According to documents obtained by CNBC, the internet giant recently convened a “internal summit” during which corporate executives and employees introduced Project Ellman. According to the article, the AI will provide a “bird’s-eye view” of someone’s life by retrieving files from your Google Account and analyzing written biographies and surrounding information. This approach entails searching through information files to identify key occasions. Google personnel said Project Ellman could figure out when a user was born, who their parents are, and so on and whether or not they have any siblings.

It doesn’t end there because it appears to be capable of highlighting chapters in your life such as the years you spent in college or living in a specific city. Ellman can even figure out what you consume. If you upload a lot of photographs of pizza and pasta, for example, the AI can deduce that you like Italian food. The technology isn’t limited to one individual either; it can recognize friends and family, as well as social events you’ve attended.

Based on the report’s description alone, Project Ellman sounds quite similar to Google Photos’ Memories, albeit on a far larger scale.

Chatbot for personal use

According to CNBC, the presentation followed with a demonstration of Ellman Chat, which is similar to ChatGPT but has the potential to “answer previously impossible questions.” According to the instances, the questions aren’t necessarily impossible; they’re just difficult, especially if you’re forgetful. For example, you can ask the chatbot when your brother last visited or for destination suggestions based on the photos you provide.

Then we get to what could be one of Project Ellman’s hidden goals. By analyzing the screenshots people share, the technology can anticipate everything from what things you might buy to your interests and future vacation plans. The presenters also proposed that it can learn which websites you visit.

All about you, according to the analysis

We don’t think we need to tell you how disturbing this all sounds. We’re talking about an AI sifting through your files, looking for any and every data it can find. Where is all that data going?

Gemini, Google’s new large language model (LLM), is said to be the model that will power Project Ellman since it is multimodal, or can accept inputs other than text. To stay current, generative AIs require a steady stream of content. Google appears to be crossing privacy lines in search of additional data to feed Gemini and keep it growing.

Ellman, of course, has no assurance of ever seeing the light of day. According to a Google representative, this is merely “early internal exploration.” If a release is planned, engineers will take the time to ensure it is useful to consumers while keeping user privacy in mind.

Please take this statement with a grain of salt. Despite its ostensibly best efforts, the corporation has a long history of privacy violations. The company gets into a lot of trouble for it. Look at the Wikipedia article on the subject; it’s massive.

Hopefully, all of this is overstated, and the tech behemoth does not unveil a digital vacuum cleaner capable of vacuuming up everything.