Insights
Research notes, documented experiments, and findings from the field. These articles record what the CogniCuisine methodology looks like in practice — including what goes wrong, what AI gets right, and where the cook always has the final word.
Process Documentation
How a Recipe Gets Built the CogniCuisine Way
Labneh Pancakes with Sour Cherry Compote & Za’atar Butter
Claude drafted it. Qwen caught three structural chemistry failures. The stove corrected what both missed. The full documented account of how an original recipe moves from AI concept through chemistry analysis to a tested, finished dish — and what the methodology looks like when the AI under scrutiny is the same one that wrote the first draft.
Read the Full Account →AI Platform Research
Why I Can’t Find My Own Recipes: The Search Problem in AI Cooking
A systematic comparison of seven major AI platforms
In 2026, AI chatbots cannot do what Gmail could do in 2004: let you search your own content. This matters deeply for anyone using AI for serious iterative work. A documented test of seven platforms reveals a consistent gap between what AI tools promise and what they actually deliver when you need to find something you know is there.
Read the Research →