This may have been pointed out elsewhere, but Ramanujan's statement was not that 1729 was the smallest number that satisfied the sum of two integer cubes; he maintained it was the smallest which could thus be described by two different pairs of integers. Therefore the 91 example does not apply.
This may have been pointed out elsewhere, but Ramanujan's statement was not that 1729 was the smallest number that satisfied the sum of two integer cubes; he maintained it was the smallest which could thus be described by two different pairs of integers. Therefore the 91 example does not apply.