In the Weights measures your digital existence in AI parameters

The site functions by prompting models including Gemini, Claude, Llama, and several versions of GPT to generate descriptions for specific individuals. It then clusters these responses to assign a strength score, reflecting how well an AI recalls a subject without relying on external web search. For the developers, the project highlights a shift in how we perceive digital footprint: as traffic migrates from traditional search engines toward LLMs, our relevance is increasingly defined by the floating-point numbers within an AI brain.

Early results show a shifting leaderboard, with figures like Macaulay Culkin currently holding top-tier scores. The platform also exposes the inconsistencies of these models, occasionally highlighting hallucinations where AI struggles to distinguish between individuals with similar names. While critics argue the project is simply a sophisticated way of asking chatbots for anecdotes, the site’s retro, Nintendo-inspired design and competitive scoring system have driven significant engagement. Dimson intends to expand the tool to analyze model bias and investigate why certain individuals gain representation in training data while others remain absent.