But he was making the point that there are valid human reasons for zero-based indexing - e.g. keeping the lower bound within the natural numbers; having the upper bound be the number of elements in the preceding sequence, etc. The "way computers think" doesn't enter the discussion.