The story in three lines. Beto notices that a mop works because its strands are floppy. Margit draws a protein with a faint pencil tail and Kemi's photograph blurs in the same place. Together they show the tail moves, and that a moving tail can catch a partner.

One thing to do at the table. Take a photo of a running pet, a waving flag or a spinning wheel, and one of something still. Ask: what does the blur tell you? Then draw something and mark what you are unsure of in pencil.

The grown-up layer. Many proteins have regions that never settle into one shape. For a long time these were treated as "missing" from structures. We now know many of them do real work: they stretch out, touch partners and let go. When a prediction tool is unsure about a region, that low confidence often means the region is floppy, not that the tool failed. "Not sure yet" is an honest answer.

For the ride home. Ask: why did Margit stop Beto from rubbing out the pencil?