A classic exchange between Tullock (1973) and Parks (1975) debated whether universities should concentrate research support on senior or junior faculty. Both made their arguments informally. We formalise the two positions and then ask which conclusions survive when the scarce input is AI compute rather than secretarial help. Under complementary inputs, aggregate-output maximisation favours juniors, while minimum-output objectives favour seniors. With AI modelled as a more substitutable input, greater substitutability alone does not reverse the junior-favouring allocation when faculty differ only in career incentives. Only the perfect substitutes limit removes the junior advantage by making the department indifferent. Reversal requires that senior faculty also possess a human-capital advantage. Sufficiently high substitutability can then shift the allocation towards seniors. A two-period extension shows why the dynamic case for juniors depends on additional primitives such as longer career horizons or steeper AI learning curves. Two appendices establish that neither the two-researcher representation nor quadratic effort costs drives the central invariance result.
