Mark Zuckerberg Raises Concerns After Confirming Meta’s Use Of Instagram and Facebook Data For Training Purposes

Meta’s CEO has raised serious concerns after speaking about the company’s future plans regarding user data from Instagram and Facebook for training purposes.

He says Meta has a plethora of data that is even more than that observed for training ChatGPT in the past. Therefore, the organization would be making use of this data for AI system training purposes.

The goal of the tech giant is to make use of both posts as well as comments for training purposes of chatbots and that’s what’s raising eyebrows with the masses in terms of toxicity as well as privacy.

The news was first reported by media giant Bloomberg and it was announced right after Zuckerberg detailed about launching the firm’s newest earnings report.

For those who might not be aware, Meta’s Facebook app is a major deal. The figure for users continues to grow as we speak and as per the latest stats of Meta, the firm is said to be on a roll.

But the continuous growth does not have Mark Zuckerberg celebrating just yet. He hopes to take advantage of this by making use of Facebook and Instagram data to produce content that’s not only unique but powerful.

The CEO hopes the next integral part of their lineup would be using the data and feedback to the best of their abilities for a large number of products. We’re seeing plenty of images get shared on these apps but they do estimate it would be greater than the usual datasets that people put out in terms of public texts and posts across a wide range of services.

To have data that surpasses that witnessed over a wide array of text utilized for ChatGPT training purposes means the firm might be in a state to create a bigger and smarter chatbot than what’s on demand right now.

As noted by Bloomberg, Meta might be in an advantageous position for two leading reasons. For starters, it has to do with the large volume of data at stake. Secondly, the fact that most of the data is conversational means it can be used to really give the chatbot a greater edge as compared to all others.

But there are two leading red flags at stake over here. For instance, Meta is training AI on posts deemed to be personal to users. Meanwhile, chats amongst friends or loved ones taking place between friends across the comments section is another point worth mentioning that has raised serious alarms of privacy being compromised.

Let’s not forget how the volume of data arising in the comments section is actually toxic on a different level altogether. There are plenty of thoughtful users giving out insights here as well but you cannot forget how it ends up in a wide array of personal attacks and serious insults that could be deemed as worrisome and racist, not to mention sexist.

Photo: Digital Information World - AIgen/HumanEdited

Read next: OpenAI To Soon Add Watermarks To Pictures Generated Using DALL-E 3
Previous Post Next Post