Dropped a draft outline of an article format into Use AI - got image interpretation from claude, then expanded on it with gpto. That multimodal input saved me 2 hours of manual translation of images to text. From visual to draft within one environment is a massive productivity gain.