Looks Good to Me: On AI Sycophancy, Context Loss, and Inverted Baselines

AI promises to make design faster. More efficient. More productive. But what actually breaks when it enters the process? "Looks Good to Me" examines nine structural problems beneath the productivity pitch. Tools that only agree with you. Context that decays with every exchange. Edges and outliers going extinct. Baselines that invert until broken becomes normal. Not future speculation. Present observation. Through sharp, sardonic essays, product designer Tanya Donska documents what actually happens when AI enters the design process—not what's supposed to happen, but what does. From sycophancy to model collapse, from estimation failures to inverted bus factors, these essays trace the real costs hidden under the efficiency promises. For designers, developers, product managers, and anyone working with or thinking critically about AI tools.