UC is supported by artificial intelligence, but it is not driven by it. The responses you see on UC are based on my own research and analysis, developed over many years of work. AI serves only as a tool to locate, organize, and articulate that knowledge more precisely.
Here is how it works:
I upload my research-based content to OpenAI, where it is converted into what are called vectors—mathematical representations of meaning. These vectors are then sent to Pinecone, a company that indexes and securely stores them for fast retrieval.
When someone enters a query in UC’s search box, OpenAI interprets the question and converts it into another vector. That vector is sent to Pinecone to check if it matches with any of my stored vectors.
If relevant matches are found, Pinecone sends them back to OpenAI, which then retrieves the related text from UC. OpenAI helps express that content in clear, natural language. What you see as the final UC response is written in understandable human language, but the knowledge it contains originates from my own analysis.
The essential point is this:
UC responses come almost entirely from my research-based materials. I allow OpenAI to enrich a response with information from its broader knowledge base only if and when that addition genuinely deepens or clarifies my content. I can also command OpenAI to use only my data and exclude any external sources.
Based on extensive testing, I consistently see only my own analytical voice in UC’s responses. Because I know my writing and reasoning well, I can immediately detect any content that does not come from my research. So far, none has appeared.
UC’s AI is not a gateway to AI-engines such as ChatGPT or Perplexity. It is a carefully designed, well-controlled system that makes my work easier to explore and understand. I will continue to monitor OpenAI’s behavior closely to ensure that my original insights, developed over more than 25 years in Germany, remain protected and represented accurately.