Picture a string as a row of numbered tiles, one character per tile, like the letter tiles in a word game. The string 'cat' is three tiles: c at position 0, a at 1, t at 2. Spaces, digits and punctuation take tiles too, so 'New York' has 8 tiles, and len('New York') is 8. The last tile is always at position len - 1, here 7.
Operations work on the tiles. + lays two rows end to end: 'Hello, ' + 'Maria' gives 'Hello, Maria'. A slice copies a stretch of tiles: 'Maria'[0:3] is 'Mar', starting at position 0 and stopping before 3. A search asks whether a shorter row appears inside: '@' in 'jo@mail.com' is True. In Python and Java a string cannot be edited tile by tile, so fixing a lowercase name means building a new string: 'M' + name[1:] turns 'maria' into 'Maria'.
The trap is digits. '42' is a string of two characters, the tile 4 and the tile 2, not the number forty-two. So '42' + '8' gives '428', and '42' + 8 is an error in Python. When text holds a quantity you want to calculate with, convert it first: int('42') + 8 is 50. When digits are a label, like a locker code 007, keep them as a string so the leading zeros survive.
Think of it like this. Beads on a necklace, one letter per bead and numbered from 0, behave like a string: you can count the beads, read any one by its number, clip two necklaces together or copy a stretch. The picture fails when you try to swap a single bead; in Python and Java you must thread a whole new necklace instead.
Core idea. Strings are sequences — you can access individual characters by index, search for substrings, and combine strings.