Weekly results

Each week two AI models write the same summaries and readers blind-pick the better one. Here's who practitioners actually preferred, week by week — a result you won't find anywhere else.