Colin Redemer, in his “The Return of the Fellowship,” proposes vastly increasing the number of fellowships—tailored programs of study—as a solution to the problems of higher education. Higher education is an unreformable morass, doomed by the challenge posed by artificial intelligence (AI). Sprightly, unbureaucratic fellowships such as the Claremont Institute, Redemer argues, offer an easier and more effective path for higher education reform.

Redemer’s fellowships are fine creatures, and they surely deserve support. But they serve boutique purposes and cannot be adapted to the mass education and credentialing functions of the universities without becoming universities—with all their attendant bureaucracies and dysfunctions. Redemer is acute about the problems afflicting the university and the virtues of fellowships, but he does not convince in his argument that the latter is a solution to the former.

The modern university is supposed to provide mass education and mass credentialing. Education is supposed to be general, educating students to pursue any career in the country. Credentialing is also meant to be general, a way for employers who do not personally know their students or their recommenders to hire a new employee with some guarantee that they will be halfway decent at their jobs. The modern mass universities aren’t supposed to be utopias of scholarly and teaching excellence—though they should aspire to those ideals. They are supposed to be guarantees of moderately rigorous all-purpose adequacy.

This function is opposite to that of Redemer’s preferred fellowships. These are meant to provide a personal seal of approval for ideologically formed cadres, steering them into leadership roles in civil society and government. They are meant to provide a form of redress to the ideological capture of mass education, but they are not themselves engaged in mass education. Fundamentally, they still depend on the mainstream K-12 and university system. The fellowships assume basic competence in the awardees, from literacy to research skills, and they employ Ph.Ds to teach. Redemer received his from the University of Aberdeen. Fellowships, as they now stand, are fundamentally dependent upon the existing university system.

Fellowships, to provide a true challenge to the universities, would have to be provided to create a teaching cadre. The fellowships would need to be expanded to provide teachers across a wide range of subjects. The fellowships would have to create a system of mutual recognition. The fellowships, as their predecessors nigh a millennium ago, would have to become universities. There might be some temporary benefit in creating new universities without the dysfunctions of the old. There probably would not be enough benefit to make the experiment worthwhile.

Depoliticized universities evolved from fellowships and would still face the AI challenge that Redemer so acutely analyzes. The AI challenge, at the least, is that most students seem prepared to cheat with AI for any task where they cannot be directly proctored. We, therefore, are evolving toward a regime in which students receive only classroom instruction, do no homework, and never acquire the habits of independent reading, independent writing, and sustained independent initiative and effort necessary to maintain the modern economy and government, much less to innovate and improve them. We will depend utterly on AI since our children will have become lazy semi-literates incapable of independent thought or work.

This is a terrible challenge. But Redemer’s fellowships do not offer a solution to it. Claremont’s fellowships are excellent at educating a limited number of fellows who already have received a decent college education to know something of the basics of America’s principles. They never claimed to offer a solution to mass cheating, mass illiteracy, and mass unwillingness to do homework, and they are not well-suited to provide one.

Not that the universities offer a better solution. Redemer accurately states that they are corrupt labyrinths. I have argued elsewhere that it is still worth fighting to reclaim the university, cantonment by cantonment—but Redemer’s judgment that it is not worth the fight is a powerful rejoinder. I think it is a necessary fight, however difficult, but Redemer may be proven right. Still, even if he is right that the universities are not worth the fight, I don’t think his fellowship alternative can succeed.

I don’t have a perfect solution. Part of the answer lies in creating a network of K-12 schools that teach students the basic habits of honor, honesty, reading, writing, initiative, and hard work. The Classical education schools offer our best current hope of providing that network for the nation. Some part lies in creating effective standardized assessments—the Classic Learning Test, ideally expanded to include equivalents of the GRE and the SAT. Some part lies in discerning between what is needed to educate good engineers to work on the latest projects of Peter Thiel and Elon Musk and what is needed to create leaders imbued with the ideals of Western civilization and able to preserve our republic.

All these parts of the solution might still fail to meet the challenge of AI. The scope of AI’s challenge seems to increase by the week, so we should not search too quickly for comprehensive solutions. What I proffer I think is a step in the right direction—but I offer that judgment very tentatively. As for Redemer’s fellowships, they will not help us restructure the university or meet the AI challenge. But they are still good ideas and worth supporting. They will help us greatly, even if they are not panacaeas.

Follow David Randall on X.