• by Laurie Sullivan , Yesterday

Google has been moving its Gemini artificial intelligence (AI) models into every part of its ecosystem.

On Tuesday, the company provided a review of three new models it delivered in July.

Google then announced in an email to users the closure of Google Assistant -- which runs on Android for phones, tablets, Wear OS smartwatches, headphones, and phone-projected Android Auto -- on September 4.

Google will replace the Android assistant for those devices by integrating Gemini -- its next AI-powered assistant -- deeper into the platforms' systems. Vehicles with Google built-in will keep Google Assistant, for now.

The change does not immediately apply to Google Home or Google TV.

A post on Google's support page also describes what to expect. 

"Google is officially transitioning from the deterministic, micro-task architecture of Google Assistant to the conversational, multimodal capabilities of Google Gemini," Aswin Jagadeesh, a Google product expert, wrote in a Google post on the support page.

Jagadeesh called this change "a major ecosystem shift" that prompted questions, so he put together a timeline for the changes. 

The move gives advertisers and media buyers reaching consumers on Android a smooth transition into serving actual real-time content based on personal context, and eventually, when the rollout is complete, it will allow the ad to serve up to consumers across mobile, wearables, and in-car systems.

Jagadeesh wrote that while the transition brings massive upgrades announced at the latest Google I/O, the upcoming "Gemini Spark" agent system will allow phones and home devices to naturally understand context, proactively manage complex schedules, and coordinate directly with apps on the users' behalf without requiring hyper-specific "wake-word" phrasing such as "hello Google."

Google wrote in a post in May that the shift to the always-on infrastructure like the Google Gemini Spark agent architecture alters the way
brands reach consumers. Rather than bidding on user search terms, advertisers must now optimize for autonomous AI agents acting as gatekeepers and transaction-handlers. 

For example, in the May post, Google began introducing AI agents for specific tasks. "Daily Brief," an agent that provides a personalized morning reading list that’s designed to be your first stop every day.

Built on the success of the Google Labs experiment "CC," Daily Brief pulls information through Gemini from across a user's connected apps to gather updates from the user's Gmail inbox, tracks upcoming events from their Calendar and compiles follow-up details into a brief notice to summarize daily activities. It also suggests immediate next steps. 

Google did not mention how brands might interact with Daily Brief in the future, but think of ads that match the user’s exact daily schedule, location, and inbox context. Promotions will begin to look more natural like "next step" suggestions in the briefing. Sponsored items could appear precisely when a user is planning their day.

This could allow the concept of zero-clicks to dominate organic optimization where users get answers directly in the agent, completely bypassing brand and publisher websites. It would mean fewer ad impressions and lower traffic rates as agents summarize data, something the advertising industry has been grabbling with for a while. 

Tuesday's announcements related to discontinuing Android and moving to Gemini continue to transition the advertising industry into using real-time user context to eliminate possible privacy concerns and improve on actual personal content.