That one was really surprising! At first it looks like complete guesswork, but then it turn out that you only have to infer the square at r2c1 (which you can easily do using the 1 in row 5 and the numbers from column 2) ... and the entire thing follows completely! Brilliant!
What I meant was this: If you try to start the 6 in column 1 from any other square than the second, you'll get contradicitons pretty quickly ... mostly with the numbers from column 2.