The rise of AI-powered search is forcing a seismic shift in how brands achieve visibility online. In a discussion on Practical AI, Liam Dunne and Ben Moore of Discovered Labs outlined how large language models are reshaping the search landscape, moving away from traditional keyword-based SEO toward a new visibility paradigm. When an LLM generates a search answer, factors like retrieval, citations, representation in model weights, and consensus all play roles in determining which sources appear—a fundamentally different game than the link-building and keyword optimization that governed Google search. The strategic implications are immediate and complex. Brands can no longer rely solely on conventional SEO tactics; instead, they must think about how to become visible within language models themselves. Dunne and Moore highlighted unconventional strategies like building a Reddit presence and optimizing for "query fan-out"—ensuring a brand's content surfaces across multiple related searches, not just exact-match queries. What companies say about themselves matters less than how the broader web represents them in the training data that powers these models. Looking further ahead, the conversation turned to AI agents that will do more than passively discover websites—they'll interact with them and take actions on behalf of users. This evolution suggests the next frontier of competition won't be about appearing in search results, but about being the system of record that agents trust enough to transact through.