I recently got access to this new capability, and let me tell you, it’s a game-changer né?. This year, a security researcher found that ChatGPT could be manipulated to implant “fake” memories, leading to potential data theft.
[See the featured image below]
né?. In the Gemini user interface Google illustrates how memory prompts can come in handy such as advising to “Use simple language and avoid jargon” or indicating coding limitations like “I can only code in JavaScript.” While the feature currently supports only English prompts users have the flexibility to disable it at any time although stored memories will persist until manually cleared out.
It’s worth mentioning that Gemini memories aren’t used for model training purposes. For now it’s only available on the web client and hasn’t made its way to the Gemini apps on iOS and Android né?. Similar to ChatGPT Gemini’s memory function adds a whole new dimension to conversations by keeping track of context. For instance, if you mention your favorite foods, the next time you seek restaurant recommendations, Gemini might just surprise you with personalized suggestions tailored to your taste buds.
As per reports from 9to5Google, this memory feature is currently exclusive to users subscribed to Google’s $20-per-month Google One AI Premium plan. So, guess what? Google’s Gemini chatbot has leveled up! Now, it’s equipped with a memory feature that can store details about your life, work, and personal preferences. A Google spokesperson reassured TechCrunch via email that “Your saved information is never shared or used for model training.” However, just like with any memory feature, there are potential risks if not adequately protected né?
