This question comes up so often that I have now added a comment to the "known errors" page on the web. (Someone asked the same question after lecture too).
You are right, if each light colored mass was one type of atom (say sodium), and each dark colored mass were a different type of atom (say chlorine) then it would be hard to understand why the bond to the left and the bond to the right would be any different. However, if each light colored mass represented a complicated *non-spherically symmetric* molecule, then there is no reason the left and right sides of the molecule have to at all behave the same.
Does this make more sense?
More interesting details you don't have to know: Even in cases where you are considering simply two types of single atoms forming a chain, call them A and B, where you might expect the two bonds to be equivalent, occassionally one gets so called "dimerization" where the two atoms form a "Dimer" molecule (A-B) which is tighly bound, and then these dimers line up only weakly bound together, so you get a chain like
And here it is clear that the bonds do alternate. Although this may look unnatural, it can happen that the above configuration is lower energy than equally spaced
In fact, this alternating spacing between atoms can even happen just for a chain of a single type of atom --- and it is even *expected* to happen for monovalent atoms. This effect was discovered by Rudolf Peierls who was the head of the theory department at Oxford for many years.