Much of what we see online — from videos to news articles to our friends’ vacation photos — is determined by algorithmic systems. Given the vast influence of these algorithms, it is important to collaboratively design guidelines and principles for how content should be chosen and ordered on AI-driven platforms.
The Content Targeting and Ranking Workstream seeks to address the open challenges of our algorithmically mediated information ecosystem, asking question such as:
- How can we create metrics that capture the effect of AI systems on human lives and values?
- Given a goal like the promotion of credible content, how can we modify the recommender systems powering online platforms to align with these goals?
Ultimately, this work hopes to translate shared goals into technical approaches that are effective at scale.