After ChatGPT’s Outage, OpenAI Reveals Memory Feature to Collect Data

OpenAI is testing a new feature, dubbed Memory, after ChatGPT’s outage to allow the generative tool to remember conversations, leading to more personalized interactions, according to the generative AI company.

Why ChatGPT Getting a Refresh?

A refresh is needed since ChatGPT operates on a single-interaction basis, processing each prompt or question autonomously. Now, with the memory feature, it can recall past details, including user preferences, or previously discussed topics, or even any information mentioned in earlier conversation segments. ChatGPT aims to create a more natural conversational experience to avoid repetitive responses and previous interactions.

On top of that and based on the major outage that occurred on the 14th of February, when users experienced a complete blackout from 8:00 PM IST until around 10:30 PM IST. Not done yet, it blacked out again at 2:06 AM.

_{Image: Down Detector}

ChatGPT users expressed on X (FKA Twitter) their discontent with the generative tool’s outage, with one user humorously stating that the end is near.

Hours following the disturbance in its tool’s services, OpenAI announced that it would be deploying the new memory feature to ChatGPT. This feature will help users save conversations, information, and discussions, which can be accessed even in the event of a complete outage.

The AI company also stated in its release that “it was taking steps to ensure ChatGPT won’t automatically log what it calls ‘sensitive information,’ such as information about your health,” according to Business Insider.

ChatGPT Outage Enables Further Data Collection

Some users have found themselves questioning the timing of the outage and the prompt deployment of the memory feature. Followed by OpenAI’s announcement that all saved chats through the memory feature will enable OpenAI to use all saved chats in future GPT models’ training.

Linking this announcement with the Archive feature announced in December 2023, it does not seem that unlikely that this could be another attempt by OpenAI to validate its access to users’ chats to train AI models.

While this has yet to be confirmed, we will leave it to the future to endorse or debunk these claims.

For now, experts highlight the need for users to be aware of what they type, as it could be used for future AI training. Although, as mentioned before, the company did indeed state that sensitive information won’t be stored, like health information, but for some users, certain chats between them and their subscribed GPT may be considered confidential.

Currently, OpenAI is addressing privacy concerns, but this also calls for users’ attention to the importance of ethical implications for AI collecting users’ data and to what extent it is being used by the tool’s parent company.

The advancements in Large Language Models (LLMs) and their ability now to engage in more complex and interactive responses are a clear representation of technology’s direction.

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.