# Why does the summation of these indicator variables start from i<j?

I'm currently reading through the eighth edition of A First Course in PROBABILITY, by Sheldon Ross. The section I'm reading is "Momens of the number of events that occur", and I understand everything below except for in the case of the summation of $$I_iI_j$$, why are they starting from $$i?

If the summation were over all $$i\ne j$$, then every pair would be counted twice.