Meta uses AI to make Instagram even more addictive

The data was presented by Mark Zuckerberg, co-founder and CEO of the company, during the release of the company's most recent financial results.

According to the executive, the advance is mainly linked to changes made to the feed and Reels recommendation system. The technology began to analyze each user's behavior in more depth to predict which publications are most likely to arouse interest.

Recommendations are trained with a large language model fueled by photos, texts and videos openly published on Instagram. With this, the system can identify not only the subjects present in each content, but also characteristics such as language, style and tone.

Zuckerberg stated that the platform also began to prioritize more recent content and offer tools that allow users to interfere more directly with what appears on the screen.

In Reels, the new system combines greater processing speed with an architecture capable of considering a more detailed history of browsing, likes, shares and other interactions.

According to Meta's CEO, the changes resulted in an increase of 15 basis points in sessions held within Instagram, in addition to boosting the sharing of publications and the time spent on the application.

The movement shows that Meta's billion-dollar investments in artificial intelligence go beyond the creation of new resources. The company also wants to use technology to reformulate content distribution and make recommendations increasingly personalized.

For creators, the promise is to expand the reach of publications. For Meta, the objective is to keep users connected for longer and encourage more interactions within the social network.

Slow Internet? These are ways to improve your online meetings

If you don't have the option of changing your router to improve the quality of your Wi-Fi network, there are certain measures you can take to improve the image and sound of your video calls.