This project started as an essay about journalism’s crisis in the age of AI — how tools like ChatGPT and Perplexity are reshaping the web’s economics and our relationship to information. AI answers feel convenient, but every time they summarize instead of linking, fewer people visit the publishers who did the reporting in the first place. Traffic — the lifeblood of journalism — drops, and so does the funding for new, factual, human-generated work.

Meanwhile, the user is asked to simply trust that the AI’s answer comes from reliable sources — whether through its training data or a quick search — without ever really knowing whose words, judgments, or biases are behind it. Every piece of writing has a perspective, but today’s AI systems tend to blur those perspectives together until the origins of information disappear entirely.

I wanted to see if there was a different way to design AI search. What if the AI answer didn’t replace the web page, but revealed it? What if you could literally see and interact with the source while you read the answer — not as a citation buried at the bottom, but as part of the experience itself?

That’s what Embedded explores. Every part of the AI’s response is tied to one real source — shown, when possible, in an embedded iframe beside the text. Hover over a sentence and you’ll see the exact quote it came from. The interface encourages you to go to the source itself, to explore rather than consume. It’s more a design experiment than a finished product (iframes have plenty of limitations), but I think it points toward a healthier way to merge AI and the open web — one that keeps trust, transparency, and context at the center.

Under the hood, Embedded uses LinkUp for its high-quality search API and reactbits.dev for some fantastic open-source UI components. The LLM used is GPT-4o.

At its core, this project is about rethinking how AI and human information coexist. Instead of collapsing the web into a single opaque “answer,” Embedded tries to make AI a guide — one that still points you back to the people who did the work.

You can see more of my work on my website, alexduncangames.com.

© 2025 Alex Duncan.