ThesisSearch.App

Thesis Search reimagines The New School’s thesis archive to match today’s radically transformed expectations around how we access and interact with information. We’re living through a historic inflection point where tools like ChatGPT and Perplexity are redefining search itself not just the interface, but the very nature of how knowledge is retrieved and presented. The old “ten blue links” model is giving way to direct, conversational answers powered by large language models (LLMs). In this new world, people expect instant, context‑aware responses—and institutions that fail to adapt risk becoming irrelevant.

This shift is already hitting the biggest players. Even Google is being forced to re‑evaluate its core product in real time. A recent Forbes article notes that roughly 60 percent of Google searches now end with no clicks at all. AI‑native platforms are stealing share, and Google is under pressure to weave Gemini more deeply into its experience. These are no longer marginal experiments—they’re existential questions.

That makes right now the perfect moment for a project like this. Academic archives, including Parsons’ own, are stuck in the past: clunky keyword boxes, buried content, and high friction. Yet the underlying documents—student theses—are often the most innovative work the school produces. Thesis Search uses modern Retrieval‑Augmented Generation (RAG), semantic embeddings, and a chat interface to bring those works back to life. Instead of just storing knowledge, it helps students, researchers, and faculty actually use it.

This isn’t only about efficiency; it’s about institutional integrity. Parsons says it is “designing a better world,” and I believe most students take that seriously. It’s baked into our classes, the questions we’re asked to explore, and the way we’re pushed to think critically about the systems we live in. Every thesis—no matter the topic or medium—is an attempt, however small, to make something better: to challenge assumptions, tell new stories, explore emerging tech, or build something useful. We pour hundreds of hours into these projects, often working past our limits, driven by that shared sense of purpose.

Here’s the contradiction: the archive meant to preserve all this work is effectively broken. It buries student ideas under layers of friction and outdated design that no longer reflect how people search and learn. If we are serious about “designing a better world,” we have to start by surfacing the intellectual labor already happening inside our own walls.

Thesis Search is my attempt to close that gap. It isn’t perfect, but it proves that thoughtful design plus the right AI tools can turn a static graveyard into a living system—one that mirrors the way people now search, question, and discover. It is more than a retrieval tool; it’s a statement that student work still matters long after the diploma is framed.