The prompt read: "You are helping me import context from one AI assistant to another. Your job is to
go through our past conversations and sum up what you know about me."
"In the output, please avoid using any first-person pronouns (I, my, me, mine) and any second-person
pronouns (you, your, yours). Instead, refer to the individual you have learned about as "the user" or use neutral phrasing."
"Preserve the user's words verbatim where possible, especially for
instructions and preferences."
Google Gemini's new process to import memories and chats from other chatbots eliminates the need to manually rebuild years of personal context,
preferences, and ongoing conversations.
It also addresses the process of turning over user context from within a closed loop into a portable asset, drastically shifting the competitive
dynamics of the consumer AI market.
For advertisers, this feature accelerates Google's ability to feed high-intent and very personal user data into its advertising ecosystem, bypassing signals
lost to privacy regulations like App Tracking Transparency (ATT), Apple's user privacy framework, or the deprecation of third-party cookies.
The alert
Mandese received from Google Gemini outlines a process known as context migration or memory porting between two AI systems.
It acts as direct command instructions to analyze historical
data of past interactions and compress that information into a portable summary.
While Gemini conversations are protected by privacy policies, the effects of a user consolidating their entire
digital life into Google's ecosystem create massive advantages for brands.
A quick search on Google AI Mode surfaced these benefits: capitalizes on zero-party data, charges Performance Max
(PMax) campaigns, and bypasses "Signal Sandbox" required by privacy regulations, for starters.
"When a user uploads a 5GB chat history filled with their career goals, health journeys, shopping
preferences, and coding projects, Google's semantic models can map a highly sophisticated, anonymized intent profile that powers hyper-targeted audience segments across Search, YouTube, and Maps,"
according to Google AI Mode.
Consumer trust could become Google's biggest hurdle to the success of this feature once the wall comes down.
In this case, the consumer allowed it to
happen, giving up their data. Once users suspect that importing their private, intimate chat histories from ChatGPT or Claude will result in them seeing hyper-specific ads on YouTube the next day,
this could trigger a privacy backlash.