A year ago, we got a new client who had never heard of us until an AI recommended our work. That moment was exciting. It felt like scale. Like something out there was telling our story for us. But then came the realization: we weren’t in control of that story.
Today, AI systems are constantly ingesting, structuring, and reshaping the content we publish. Our websites, once designed for human audiences, are now quietly becoming training data for machines that interpret, summarize, and speak on our behalf. This is happening whether we like it or not.
In this keynote, we’ll explore how content models, metadata, and information architecture are shaping AI systems in ways most teams never intended. We’ll look at what happens when content is treated as disposable output instead of durable knowledge, and why that shift now carries real consequences.
The question isn’t whether AI is learning from your website. It’s whether you’re comfortable with what it’s learning.
Jim is an Engineer Manager at Four Kitchens who specializes in creating web architectures – translating application requirements into useful, discrete, and reusable components. Jim uses web frameworks, including Drupal to support all of his nerdy hobbies, including planning retro-arcade projects; sharing photos of outdoor adventures; and organizing Philadelphia-area summer camps.
