Source: Yahoo

After years of watching content that keeps them hooked on social media platforms, users of Facebook and Instagram will finally understand how these apps keep them locked. This is because Meta has finally published an explanation as to how its AI models select content to recommend to users on different platforms.

Behind Meta’s AI Recommendations

Lately, Instagram seems to know what a user wants and recommends it to them. Sometimes, a user could mention a product to a friend, and the next time they check the app, the product will be shown on their feed. The same has been the case for Facebook for years where users have received news on their feed that matches their interest.

However, many users have been concerned that the company is listening to them due to how scaringly accurate the app’s recommendations have been. Unfortunately, most users do not understand how the recommendation algorithm works.

In fact, according to a Pew Research Center Survey, over 50% of all users over the age of 18 have no understanding of why Facebook chooses to show them what it does. To address these concerns and shed some light on the black box, Meta published a breakdown of the company’s social media algorithms.

Meta’s President of Global Affairs, Nick Clegg, said in a blog:

“With rapid advances taking place with powerful technologies like generative AI, it’s understandable that people are both excited by the possibilities and concerned about the risks. We believe that the best way to respond to those concerns is with openness.”

To do this, Meta released 22 “system cards” that cover the Feed, Stories, Reels, and other methods that users find and consume material on Meta’s social media platforms. Each of these cards gives comprehensive yet understandable details on how the artificial intelligence (AI) systems that power these features rank and suggest content.

In addition to this, Meta also shared information about signals, such as liking and sharing a post, and how these inputs influence a user’s recommendations. This information was made available through the company’s Transparency Center, which is where Meta shares its policies and transparency reports in order to keep users informed.

Clegg also clarified that the company would only reveal some of the signals, not all. “While we want to be transparent about how we try to keep bad content away from people’s Feeds, we also need to be careful not to disclose signals which might make it easier for people to circumvent our defenses,” he explained.

Meta on More User Control

This is not the first attempt Meta has made in letting people into the algorithm’s inner workings. In 2019, the company launched a feature dubbed “Why am I seeing this?” which helps users understand why they were being presented with certain content. The feature also lets users control what was being recommended to them.

Meta has decided to expand the use of this feature to the Instagram Reels tab and Explore, and Facebook Reels in the coming weeks after previously only using it on posts and ads. By clicking on the reel, a tab will pop up with information on previous activities that led to the specific recommendation.

Source: Meta

As a way of offering users more control over what they are presented with, Instagram added a feature to say “Not Interested” on the content they did not like. The company is now extending the feature to allow users to mark content they want to see more of as “Interested”. The feature is currently being tested on Instagram reels before release.

By being more transparent about how its system works, Meta is hoping to dissipate some of the anxiety users have built around their privacy in relation to the platforms’ algorithms.

Meta has embarked on a deep AI journey with its platforms more recently. In its Q1 earnings calls, the company revealed that more than 20% of the content on its platforms is recommended by AI.

By employing the technology, Meta reported a 24% increase in the amount of time users spend on Instagram and Facebook. As a result, the company is now working on pushing the AI influence on at least 30% of the content.

What's the Best Crypto to Buy Now?

  • B2C Listed the Top Rated Cryptocurrencies for 2023
  • Get Early Access to Presales & Private Sales
  • KYC Verified & Audited, Public Teams
  • Most Voted for Tokens on CoinSniper
  • Upcoming Listings on Exchanges, NFT Drops