News and Information

Assignment: “Freespoke’s Technology Manifesto” (May 15, 2024), “Freespoke AI Index Manifesto” (October 9, 2025), “How Freespoke 'unbiases' Biased LLMs” (December 1, 2025)

We often speak about “the news” as though it were a transparent window into the world: events happen, journalists report them, and the rest of us learn what is true. But news and information never reach us without selection and interpretation. Someone decides which events deserve attention, which sources are credible, which facts belong in the story, what context matters, and what the headline should emphasize. These decisions do not necessarily reflect dishonesty or partisan manipulation; they are unavoidable features of communicating about a complicated world. But they become more consequential in a media environment shaped by speed, competition, audience loyalty, algorithms, and the constant struggle for our attention.

The problem, then, is not solved simply by finding an “unbiased” source. Different publications may report the same facts while framing their significance differently, and our own prior commitments affect which accounts we trust, question, or ignore. At the same time, recognizing that every source has a perspective does not mean that every claim is equally reliable or that truth is merely a matter of opinion. Learning to navigate news well requires judgment: distinguishing reporting from analysis and opinion, tracing claims back to evidence, comparing how different sources cover the same event, noticing what has been emphasized or omitted, and remaining open to information that unsettles our existing views. More information does not necessarily produce more understanding—especially when we do not see the same stories, trust the same institutions, or agree about what counts as relevant evidence.

We will explore these challenges through Freespoke, a search and news platform designed to help users compare information across political perspectives. Freespoke labels publishers—not individual articles—as left, middle, or right, drawing in part on third-party media ratings, and organizes coverage so users can see different accounts of the same topic alongside one another. It also offers AI-generated overviews intended to identify areas of agreement, disagreement, and expert opinion. These features are meant to make the filtering of information more visible and to give users greater agency over what they encounter. But they also raise questions. How useful are broad political labels, and what might they obscure? Does presenting multiple perspectives help us recognize our blind spots, or can it create a false equivalence between stronger and weaker claims? What can a platform tell us about perspective that it cannot tell us about accuracy, expertise, or truth? Our goal is not simply to find the right information source, but to become more discerning about how information reaches us, how it forms us, and how we decide what—and whom—to trust.