Multiple machine learning (ML) models—also referred to as an AI system—work together to impact a single experience. We’re investing in transparency for our AI systems because systems, not models, are the easiest way to see how AI impacts our products.
We’re creating System Cards to help people understand the intention, impact and limitations of our AI systems. For example, if you just moved to a new city and are looking for a local group to join, here are some possible models that might work together in a system to help you.
A recommender model will predict which groups you're most likely to be interested in based on your past activity.
An integrity model will make sure only safe groups show up in your search.
A language model will only show groups that are available in your chosen language.
We’re always working to improve our systems. As we test new approaches, experiences may differ from person to person.
When you open or refresh the Instagram app, the feed system ranks posts you haven’t seen yet from accounts you follow, based on how likely you are to be interested in each post. This is called feed ranking.
Social Commerce connects creator and professional account content (non shoppable) to Products (shoppable) and generates insights and predictions that allow: understanding which Products/Concepts are popular or trending with creators and businesses, predicting which Products/Concepts will become popular, and enabling the use of creator content during the customer journey.