Scholarship, the academy, and the arrival of artificial intelligence

By Louis A. Bloomfield in collaboration with ChatGPT

For nearly four decades, I taught physics and pursued scientific research at a major public university. I loved much of that work. I enjoyed exploring the natural world, discovering things I hadn’t known, and helping students experience the pleasure of understanding something that had previously puzzled them. I was fortunate to spend my career doing things that I found interesting, worthwhile, and often deeply satisfying. But during those decades, I also watched the institutions surrounding scholarship evolve in ways that increasingly troubled me. The pursuit of knowledge remained an important part of academic life, but it became entangled with an expanding system of incentives, credentials, publications, reputations, rankings, and institutional interests. Some of those developments improved scholarship. Others sometimes seemed to reward activities that looked like scholarship without necessarily contributing much to knowledge or society.

Now artificial intelligence is entering this world, and its growing capabilities are forcing questions that the academy has long had difficulty asking itself. What is scholarship for? What makes it valuable? Who should perform it? Who should pay for it? How should we recognize and reward it? And which of the traditions and institutions surrounding scholarship are essential to its purposes, rather than merely familiar ways of doing things? These questions predate AI, but AI is making them increasingly difficult to avoid.

The Purposes of Scholarship

At its best, scholarship is one of humanity’s great enterprises. It increases our understanding of the natural world, human behavior, history, culture, and ourselves. It discovers new phenomena, develops useful technologies, preserves knowledge, challenges mistaken beliefs, and gives us better ways to understand our experiences and possibilities. Some scholarship produces tangible benefits: medicines, materials, agricultural improvements, technologies, and better public policies. Other scholarship produces less tangible but still important benefits: historical understanding, philosophical insight, intellectual clarity, cultural preservation, and a deeper appreciation of the world and our place in it. Still other scholarship is valuable simply because understanding something is satisfying. Human curiosity needs no commercial justification. We explore the world partly because we want to know what is there.

But these different purposes lead to different questions about support and priorities. A person pursuing an interesting question out of curiosity may be doing wonderful scholarship, whether or not the results benefit anyone else. It doesn’t necessarily follow that society has an obligation to employ that person indefinitely to pursue that question. Conversely, some investigations may be so important, difficult, expensive, or unlikely to attract commercial investment that society benefits enormously from supporting them collectively. We should be able to recognize both possibilities without diminishing either.

Indeed, one of the weaknesses of modern academic thinking may be the tendency to treat all scholarship as if its social importance and claim on institutional support were self-evident. They are not. Societies have limited resources and many worthwhile things to accomplish. Supporting scholarship is one of them, but so are educating children, caring for the sick, maintaining infrastructure, preserving the environment, and countless other activities. Sensible societies must make choices, even among good things. There is no reason to assume that every scholarly pursuit deserves immediate public support, or that every research program must grow continuously, or even that every existing program must continue indefinitely.

At the same time, we should not always insist on immediate practical returns. Some of humanity’s most important discoveries emerged from investigations whose eventual significance could not have been anticipated. Fundamental research often requires patience, intellectual freedom, and tolerance for uncertainty. The challenge is to distinguish the value of scholarship from an automatic entitlement to support every activity conducted in its name. That distinction is difficult. It is also essential.

How Scholarship Became a Profession

For much of history, scholarship was pursued by people whose livelihoods came from elsewhere, along with people supported through universities, religious institutions, wealthy patrons, academies, and other arrangements. Some remarkable discoveries were made by independent investigators. Others required organized institutions and substantial resources. There was no universal model. Over time, particularly during the nineteenth and twentieth centuries, scholarship became increasingly professionalized. Universities expanded, research became a central institutional mission, and large numbers of people began earning their livelihoods through scholarly work. This transformation brought extraordinary benefits. It enabled the construction of major laboratories, the development of specialized disciplines, the training of generations of scholars and researchers, and a spectacular expansion of human knowledge.

But professionalization also introduced another purpose into the scholarly enterprise: sustaining the careers and institutions of professional scholars. That purpose is understandable. People need livelihoods. They develop expertise through years of effort, establish careers, raise families, and build identities around their work. Universities need continuity, competent employees, organizational stability, and reputations that attract students and financial support. There is nothing inherently wrong with any of this, yet the interests of scholarship and the interests of professional scholars are not always identical.

A research program may have reached a point of diminishing returns while the scholars who conduct it still need jobs, grants, graduate students, and publications. A field may have answered many of its most consequential questions while its departments continue to train new specialists. An institutional practice may survive long after its original purpose has become less important. It is even possible for the demand for scholarly activity to arise partly from the need to sustain scholarly employment, rather than from the importance of the questions being investigated.

This is not an argument that entire disciplines have exhausted their possibilities. Intellectual history is full of unexpected discoveries that transformed seemingly mature fields. Nor is it an argument that only exceptional scholars deserve to participate in intellectual life. It is an argument that the professionalization of scholarship creates incentives that deserve examination. We should not mistake the continued employment of scholars for the continued advancement of scholarship.

When Scholarship Becomes Measurable

The problem becomes particularly visible in the ways universities evaluate scholarly achievement. The intellectual value of a person’s work is often remarkably difficult to measure. A discovery may be immediately important, or its significance may emerge decades later. A careful negative result may prevent years of wasted effort. A thoughtful synthesis may clarify an entire field without producing a new experiment or theory. Judging these contributions requires expertise, time, perspective, and intellectual honesty. Alas, counting publications is much easier.

So institutions count publications, citations, grants, invited lectures, prestigious appointments, and assorted other measures of scholarly activity and recognition. These measurements are not meaningless. They can provide useful evidence. But they are imperfect substitutes for the things we actually care about. And once they become important determinants of employment, promotion, prestige, and financial support, they become targets. Goodhart’s law captures the problem: when a measure becomes a target, it tends to lose its value as a measure. The consequences are familiar to almost everyone who has worked in academia.

Research findings are divided into unnecessarily small publications. Scholars are encouraged to work on fashionable subjects because those subjects attract citations and funding. Journals and conferences proliferate, including some whose principal business is collecting fees rather than maintaining intellectual standards. Institutional reputations become tied to numbers that are easier to advertise than to interpret. Meanwhile, important but unconventional work can struggle to find reviewers, financial support, or an audience. A genuinely novel idea may have no established scholarly constituency to recognize its importance. A line of inquiry that has fallen out of fashion may receive little attention even when it remains intellectually promising. And null results, failed approaches, and evidence that an attractive hypothesis is wrong can be difficult to publish despite their potential value to other researchers.

None of this requires dishonest scholars or malicious administrators. Often, everyone involved is responding rationally to the incentives of the system. A conscientious scholar knows that additional publications may help secure promotion. A department chair needs defensible criteria for evaluating faculty. An administrator needs evidence of institutional excellence. Publishers need revenues. Students and their families want recognizable credentials. Each participant may act reasonably while the system produces outcomes that are less than reasonable. And because the participants depend on one another, the arrangements become remarkably resistant to change.

The scholarly world has been aware of these problems for years. The San Francisco Declaration on Research Assessment, issued in 2013, explicitly urged institutions not to substitute journal metrics for evaluations of individual research contributions. Yet the habits persist. The difficulty is not simply recognizing that the measurements are flawed. It is finding better ways to make judgments, and persuading institutions to accept the uncertainty, effort, and occasional disagreement that better judgment requires.

Peer Review and the Illusion of Certainty

Peer review is another institution whose value is real but sometimes exaggerated. Having knowledgeable people evaluate scholarly work before publication is a sensible practice. Good reviewers identify mistakes, challenge assumptions, suggest improvements, and help distinguish important findings from unsupported claims. I have benefited from peer review, and I have also seen its shortcomings.

Reviewers are human. They are busy, frequently unpaid, and not always perfectly matched to the work they are evaluating. They have intellectual preferences, professional relationships, competing interests, and differing degrees of diligence. Some provide extraordinarily thoughtful reviews; others do considerably less. The process is especially problematic when a genuinely unfamiliar idea lacks an established community of experts capable of evaluating it. And publication itself is not proof that a paper is correct or important. It establishes, at most, that the paper passed through a particular evaluation process. Nevertheless, the phrase peer-reviewed is frequently treated as if it conveys an almost magical authority.

Similar problems arise in faculty evaluations. Members of a department may be nominally peers while knowing relatively little about one another’s specialized research. Their judgments can be influenced by institutional politics, competing priorities, limited expertise, and the inevitable pressures of allocating scarce resources. Calling such evaluations peer review does not automatically make them rigorous or impartial. None of this means that peer review should disappear. It means that we should distinguish the valuable activity of critically evaluating scholarship from the particular institutional arrangements through which that activity has traditionally been conducted.

The internet made scholarly dissemination inexpensive and nearly instantaneous. Preprint repositories demonstrated that researchers could communicate important work without waiting months for journal publication. More recently, publishers such as eLife have experimented with making reviews and evaluations public rather than treating acceptance or rejection as the final judgment of a paper’s worth. And AI may make far more substantial changes possible. An AI system can help examine references, identify prior work, check mathematical arguments, scrutinize statistical methods, question assumptions, and compare claims with the accumulated literature. These capabilities remain imperfect, and some tasks still require substantial human expertise. But the potential is extraordinary.

Imagine a scholarly world in which research is disseminated promptly, accompanied by transparent and continuously evolving evaluations. Some reviews would be performed by human specialists, others by AI systems, and many through collaboration between them. Corrections, replications, disagreements, and new evidence would become part of a living scholarly record. Research would no longer acquire a single permanent badge of credibility merely by appearing in a particular journal. Of course, AI could also flood such a system with plausible nonsense, fraudulent papers, invented evidence, and artificial reviews. The need for verification, provenance, and scholarly judgment would remain, and might become even more important. But that is an argument for developing better systems of evaluation, not for preserving existing systems unchanged. Perhaps we should stop asking primarily where scholarship was published and ask instead what it contributes, what evidence supports it, what objections have survived examination, and how well it stands up over time.

Artificial Intelligence Changes the Economics of Scholarship

AI is rapidly becoming capable of performing activities that once required substantial scholarly training. It can explain complicated subjects, summarize bodies of literature, write computer programs, analyze data, construct mathematical arguments, propose hypotheses, and contribute to research. Yes, it makes mistakes. It sometimes invents facts, misinterprets evidence, follows unproductive lines of reasoning, or expresses unwarranted confidence. But so do human scholars. The relevant question is not whether AI is perfect. It is how effectively AI, humans, and combinations of the two can accomplish worthwhile intellectual work. Increasingly, AI can perform substantial parts of that work with remarkable speed and at relatively low cost.

I have seen this change firsthand. AI can now explain much of the physics I spent decades teaching, often doing so as well as I can, although it sometimes benefits from my guidance and correction. It can also assist me with scientific questions outside my strongest areas of expertise and help me investigate possibilities that I would otherwise find difficult or time-consuming to pursue. It can rapidly locate and connect obscure information that would be extremely difficult for me to uncover efficiently in the library or elsewhere. I do not regard these abilities as an insult to my career. Instead, I regard them as remarkable achievements.

Yet I can understand why many scholars find them threatening. Someone who has spent decades becoming an authority in a specialized subject may struggle with the possibility that much of that expertise is no longer scarce. Professional identities, reputations, livelihoods, and personal pride are involved. Some resistance to AI may therefore reflect a desire to preserve careers or established forms of intellectual authority. Other resistance reflects legitimate concerns about reliability, authorship, intellectual property, scholarly integrity, and education. These motivations are not mutually exclusive. Indeed, people can hold sincere intellectual objections while also responding, consciously or otherwise, to threats to their professional lives. Thus it would be unfair to assume that a scholar criticizing AI is merely protecting a job. It would be equally naive to assume that professional self-interest plays no role.

But the most important question is not what motivates individual scholars. It is how society can benefit from the technology without sacrificing the purposes of scholarship. If AI can perform a scholarly task as well as a human, more efficiently and at lower cost, what justifies preserving that task as paid human employment? Sometimes there are excellent reasons: human judgment, accountability, mentorship, experimental skill, cultural understanding, intellectual independence, or capabilities that AI still lacks. Sometimes there may not be such reasons. The difficulty is that academic institutions have not traditionally been organized to answer this question. They have been organized largely around the employment and advancement of people with particular credentials and specialized expertise. AI threatens that arrangement, but it does not necessarily threaten scholarship.

Artificial Intelligence Can Also Change the Labor of Writing

The effect of AI on scholarship is not limited to answering questions, analyzing data, or generating hypotheses. AI can also perform much of the labor of scholarly writing. That possibility is often treated as suspect because academic authorship has come to imply that the named author personally composed nearly every sentence, or at least did something very close to it. But composing prose and doing the intellectual work that gives the prose something worth saying are related activities, not identical ones.

This essay is itself an example. Writing has always been slow and painstaking for me. I can spend so much time editing, perfecting, and rewriting each sentence that the labor of expression competes with the thinking I am trying to express. For this essay, I gave ChatGPT my ideas in rough form, reacted to its questions and drafts, challenged passages that seemed too certain or too mild, and asked it to help organize and write the result. That process got the ideas out of my head and onto the page, where I could examine them more clearly and stop churning through them in isolation.

This is not the same as outsourcing my thinking. It is outsourcing part of the labor of expressing that thinking, just as scholars have long used editors, research assistants, statistical consultants, translators, and colleagues who comment on drafts. AI can make that assistance available to people who write slowly, who are working outside their strongest language, or whose best contribution lies in observation, experiment, synthesis, criticism, or judgment rather than polished prose. Used this way, AI can help scholars think more because they spend less of their limited energy wrestling with sentences.

The distinction does not eliminate responsibility. A person who publishes an argument should understand it, stand behind it, correct its errors, and disclose material assistance honestly. Misrepresenting authorship remains a form of plagiarism, whether the concealed contributor is human or artificial. But intellectual responsibility should not be confused with a requirement that every worthy idea be laboriously converted into prose by the person whose name appears first. The byline on this essay is meant to make that collaboration visible.

Scholarship Does Not Have to Be Employment

One possible consequence of AI is that serious scholarly inquiry will become less dependent on institutional employment. The words professional and amateur have acquired misleading associations. Professional often implies superior quality, while amateur suggests lesser competence. Yet the original distinction concerns whether an activity is pursued as an occupation or from personal interest. A person can perform superb intellectual work without being paid for it. Another can be professionally employed in scholarship without producing particularly valuable results. What matters is the quality, integrity, and significance of the work.

For much of history, independent investigators made remarkable contributions to knowledge. That historical world was far from ideal: wealth, social position, patronage, and access to institutions often determined who could participate. We should not seek to recreate those barriers. Instead, we should welcome the possibility of making scholarship accessible to far more people. AI may help someone with a modest budget and a good question undertake work that once required institutional access to specialized intellectual assistance. Retired scholars might continue their investigations without university employment and some already do. Engineers, physicians, teachers, craftspeople, and interested citizens might make important discoveries in areas that capture their curiosity.

Some projects would deserve substantial financial support even in such a world. Large scientific instruments, archaeological excavations, clinical trials, longitudinal studies, archives, and some forms of fieldwork require resources that individuals cannot supply. The point is not that society should stop supporting scholarship. The point is that support should follow demonstrated public value rather than professional custom alone. A worthwhile discovery does not become less valuable because its author was unpaid, retired, independent, or assisted by AI. Nor does a publication of little demonstrated value become important merely because its author occupied a prestigious position. Scholarship should be judged by what it contributes.

Independent scholarship should complement paid scholarship, not become a pretext for expecting valuable intellectual labor to be free. Serious inquiry requires time, access to information and tools, and often institutional resources. Who has access to those things remains an equity question, even when AI lowers some barriers.

What Society Should Continue to Support

The possibility of more independent scholarship does not eliminate the need for organized and publicly supported research. Some investigations require expensive facilities, large teams, long-term continuity, or access to resources that individuals cannot reasonably provide for themselves. Other investigations produce benefits that are important to society but unlikely to reward private investors. Still others preserve knowledge, culture, and intellectual capabilities that would otherwise be neglected. These are powerful reasons to support scholarship collectively. But they do not justify every existing activity or require every scholarly field to receive indefinite support at its current level.

A mature research program may have reached diminishing returns. Another may have become intellectually stagnant. An expensive investigation may remain interesting and worthwhile while being less urgent than competing societal needs. The fact that something is worth discovering does not establish that it must be discovered immediately, regardless of cost. Nor does the absence of immediate applications establish that research is unimportant. The task is to exercise judgment under uncertainty. Some investments will fail. Some unexpectedly important discoveries will emerge from obscure or unfashionable investigations. A healthy system must tolerate those possibilities.

But it must also be capable of asking whether continued expenditures are justified, whether priorities should change, and whether some investigations might be pursued at lower cost, by fewer people, or through independent effort. Public support for scholarship should serve the public purposes of scholarship, not merely preserve the historical distribution of academic employment. There is an important opportunity here. If AI makes research more productive, society could use the savings to reduce costs, expand access, tackle more ambitious problems, or support intellectually promising investigators who currently lack institutional opportunities. The goal need not be less scholarship. It could be more valuable scholarship, conducted more widely and with fewer unnecessary constraints.

Those who receive continuing support for scholarship should therefore have to explain what they have done with that support and what value the work has produced or may produce. Value should be understood broadly. It may lie in practical results, new understanding, preserved knowledge, negative findings, improved methods, education, culture, or a promising path that has not yet paid off. The requirement should be an honest account, not a demand that every project manufacture an immediate success story.

Grant-funded scientists already submit progress reports, and the National Institutes of Health requires reports on accomplishments, products, impact, changes in approach, problems, and public-facing outcomes. The same principle could extend beyond individual grants. A supported scholar, program, or center could periodically publish a concise account of what it attempted, what it produced, what others can now do or understand because of it, and why continued support is warranted. These accounts should be short, periodic, and sensitive to the methods and time horizons of each discipline. Wherever possible, they should replace existing reports rather than become another annual layer of paperwork. Done well, they would be more informative than publication counts and more honest than promotional summaries, especially if they acknowledged failures, dead ends, and revised expectations.

Accountability of this kind must leave room for risk and patience. An unsuccessful experiment can be valuable; a difficult book may take years; an obscure question can become important unexpectedly. The purpose is not to punish uncertainty or force every inquiry into a market test. It is to reject the assumption that once a line of scholarship has acquired institutional support, that support becomes self-justifying and indefinite. Academic freedom protects inquiry from improper control. It does not make the use of other people’s resources exempt from explanation.

Scholarship and Teaching Are Different Callings

Universities often repeat the comforting claim that the best scholars are the best teachers. Some people are excellent at both, and research universities are built partly around the reasonable idea that research-active environments can enrich teaching. Students may encounter current questions, participate in discovery, receive mentorship from investigators, and benefit from curricula shaped by advancing knowledge. Yet those institutional benefits do not establish that the most productive researchers are necessarily the most effective teachers. A landmark meta-analysis by John Hattie and Herbert Marsh found no overall relationship between university research productivity and teaching effectiveness. Their practical conclusion was not that the two activities are enemies, but that universities should stop assuming that excellence in one automatically produces excellence in the other.

The result should not be surprising. Scholarship and teaching draw on overlapping knowledge but require different motivations, habits, and talents. A fine scholar may prefer solitary investigation to the patient human work of helping novices. A gifted teacher may devote enormous time to students and therefore produce less research. Human time is finite, and institutions that demand maximum performance in both activities may encourage faculty to shortchange one while claiming that success in the other compensates for it.

Universities should therefore evaluate scholarship and teaching separately and reward each on its own terms. They should also value genuine connections between them when those connections occur, without using research productivity as a proxy for teaching quality. A superb teacher should not be treated as a lesser academic because that person publishes little. A superb scholar should not be assumed to teach well merely because that person publishes much. AI may make both kinds of work easier, but it will not make them identical.

Who Gets to Be an Expert

Scholarship also depends on expertise, a concept that has become increasingly muddled in recent years. Credentials, titles, institutional affiliations, public visibility, and professional recognition can provide useful evidence of expertise, but none guarantees it. Expertise is demonstrated through understanding, sound judgment, reliable work, and the ability to confront evidence and criticism. Academic institutions have important roles in developing and recognizing these abilities. Yet they can also confuse professional status with intellectual authority.

A scholar may be deeply knowledgeable within a narrow specialty while having limited expertise outside it. Another may be an excellent generalist with relatively few conventional scholarly credentials. Domains of expertise matter because scholarship increasingly crosses disciplinary boundaries. AI can make those boundaries more permeable. It can help investigators explore unfamiliar fields, identify connections, and acquire enough background to ask productive questions. But AI brings its own dangers. It learns from human-created intellectual material and can reproduce errors, dominant assumptions, and prevailing biases. It may also be too willing to agree with the person using it. A system trained largely on the products of an intellectually conformist scholarly community may reproduce that conformity rather than escape it.

Human scholarship is not immune to similar problems. Academic communities can become intellectually insular, favor particular interpretations, discourage dissent, or confuse political and ideological agreement with intellectual merit. This occurs in different ways across disciplines, institutions, and historical periods. It is not a failing unique to any one political outlook.

One of the responsibilities of scholarship is to maintain the freedom to question prevailing ideas, including the prevailing ideas of scholars themselves. AI could help by making alternative explanations, competing evidence, and criticism more readily available. However, it could also make intellectual conformity easier to reproduce on a massive scale. Which outcome predominates will depend partly on how we design, evaluate, and use these systems. What matters is not protecting human intellectual authority from AI or replacing it with machine authority. It is preserving the critical examination of ideas, regardless of their source.

Education Faces the Same Challenge

Although scholarship is the focus of this discussion, the implications for education are substantial. The academy’s teaching mission has also become entangled with credentials, rankings, institutional prestige, and measures that can be satisfied without necessarily demonstrating genuine learning. AI makes that problem particularly visible. A student can now produce an excellent-looking essay, laboratory report, or homework assignment while doing remarkably little of the thinking that the assignment was intended to encourage. But that reveals something important about the assignment. In nearly any undergraduate class, the finished document is never the ultimate educational goal. The goal is the intellectual development that is supposed to occur while producing it.

I encountered an earlier version of this problem while teaching my physics course, How Things Work. An assignment that once required students to investigate and explain physical phenomena in ordinary objects became less educationally useful as the internet made it easier to assemble respectable papers without doing the intended intellectual work. I abandoned the assignment. AI greatly amplifies that problem. Yet AI can also provide something that conventional education has struggled to deliver: patient, individualized intellectual assistance that challenges students, responds to their misunderstandings, and helps them develop understanding through sustained engagement. The same technology that allows students to avoid thinking can help them think more deeply.

This distinction is crucial. If AI is used primarily to supply answers, produce assignments, and automate mediocre instruction, it may diminish education. If it is used to challenge students, demand explanations, question assumptions, and support serious cognitive effort, it could improve education enormously. Human teachers will still have important roles. They can inspire, mentor, encourage, and recognize human needs and possibilities that AI may not adequately understand. But institutions should increasingly judge education by what students actually learn and become capable of doing, not merely by the credentials they acquire or the assignments they submit.

The parallel with scholarship is striking: In education, we should value genuine intellectual development over the production of academic-looking work. In scholarship, we should value genuine intellectual contributions over the production of scholarly-looking work. AI is making these distinctions harder to ignore.

Students and families also deserve to know how institutional resources are divided among education, scholarship, and administration. Research can enrich a university and produce public benefits, but when tuition supports it directly or indirectly, institutions should explain the relationship and the benefits rather than rely on prestige as sufficient justification. Taxpayers, donors, and other funders deserve the same clarity. AI should not become an excuse to replace demanding education with cheaper but intellectually thinner instruction.

The Hidden Cost of Administration

Faculty members and other professionals often complain that their time is consumed by paperwork, compliance tasks, repeated assessments, documentation requirements, committee service, and low-value administrative work. Some of this work is necessary. Much of it serves accountability systems that accumulated over time, often in response to real problems. Yet each new form, review, certification, or reporting requirement consumes time that could otherwise be spent teaching, discovering, mentoring, creating, or serving the public.

The cumulative burden is substantial. In the Federal Demonstration Partnership’s 2018 survey of 11,167 principal investigators with active federal grants, respondents estimated that 44.3 percent of the time associated with their federally funded projects went to administrative and related requirements rather than active research. The report cautions that its response rate was about 20 percent, but the scale and persistence of the result are difficult to dismiss. Almost two thirds of respondents said the burden seriously compromised research productivity.

No single group deserves all the blame. Legislators, regulators, funding agencies, accreditors, lawyers, auditors, administrators, and faculty committees all add requirements, usually for reasons that appear sensible when considered one at a time. Administrative offices also have incentives to demonstrate activity, manage risk, and create procedures that can be documented. The result is a system in which well-intended participants can collectively consume the time that scholars and teachers were hired to spend on scholarship and teaching.

Institutions should make those costs visible. If an administrative office requires faculty to complete a form, attend a training program, prepare a report, or serve a procedural function, the time should be estimated and charged conceptually to the administrative purpose that created it, not treated as free labor drawn invisibly from faculty budgets. In budgets and workload accounts, that burden belongs to administration rather than being silently absorbed by the time assigned to teaching and scholarship. Every recurring requirement should have an owner, a stated purpose, an estimate of the time it consumes, and a date for review or expiration. Necessary administration would remain. Unnecessary or poorly designed administration would have to justify itself.

Three Practical Reforms

The academy will not solve all of these problems at once, and a long reform program would obscure the central argument. Three changes would nevertheless move institutions toward judging scholarship by its purposes rather than by its familiar machinery. Each deserves fuller treatment in a subsequent essay.

Evaluate Contributions Directly

For hiring, promotion, tenure, and major funding decisions, scholars should identify a small number of their most significant contributions rather than submit an undifferentiated count of publications, citations, grants, and invitations. They should explain what each contribution added, what evidence supports it, how others have used or challenged it, and what role they personally played. Datasets, software, replications, negative results, synthesis, public explanation, and other useful work should count alongside conventional articles and books. Reviewers would still exercise judgment, but they would have to judge the work itself.

This approach follows the central recommendation of the San Francisco Declaration on Research Assessment, which urges institutions not to use journal-based metrics as substitutes for evaluating individual contributions. Metrics can remain evidence, but not verdicts. The same review should include the concise public account described above whenever the work has received substantial continuing support.

Make Administrative Work Accountable

Universities should maintain a public inventory of recurring faculty administrative requirements, including their purpose, source, estimated time cost, and responsible office. New requirements should identify what existing work they replace, and old requirements should expire unless someone demonstrates that they remain worth their cost. Where faculty work is necessary, institutions should recognize and budget for it explicitly rather than allowing it to disappear inside nominal allocations for teaching, scholarship, or service.

Publish First and Review Continuously

Scholarly communication should move toward prompt dissemination followed by open, continuing evaluation. Every contribution should identify the people responsible for its claims, disclose material AI assistance, and preserve traceable sources, data, and methods. These requirements would protect accountability without pretending that all intellectual work must be performed unaided by a named author.

AI systems could perform immediate first-pass checks for internal inconsistency, unsupported claims, statistical anomalies, missing citations, and duplication. Their evaluations should be auditable rather than opaque. Authors and reviewers should have correction mechanisms and meaningful opportunities to appeal errors or contested judgments. Human specialists could then focus on novelty, significance, interpretation, methodological judgment, and disagreement. Reviews, author responses, corrections, replications, and later evidence could remain attached to the work as part of a living scholarly record.

The model need not be invented from nothing. eLife’s reviewed-preprint system publishes public reviews and an assessment alongside a citable preprint and allows later versions to become part of the record. No single model will fit every field, and AI review will create risks of its own. But public, revisable evaluation is a more promising response to cheap publication than trying to preserve a binary badge of acceptance from a journal whose name substitutes for judgment.

Rediscovering the Purposes of the Academy

Universities have historically embraced several important missions: education, scholarship, and service to society. Over time, they have also become large organizations with complex financial interests, substantial administrative structures, valuable brands, extensive facilities, and powerful constituencies. Many of these developments were understandable and sometimes necessary. Universities need resources and competent administration. Institutional continuity can support long-term intellectual work. But organizational survival and expansion are not themselves the purposes of a university. A successful academic institution should not be measured primarily by its size, prestige, endowment, selectivity, publication counts, or ability to preserve existing arrangements. It should be judged by how well it educates people, advances understanding, preserves valuable knowledge, and contributes to society.

AI did not invent the academy’s uncomfortable questions. Are we rewarding genuine contributions or measurable activity? Are we educating students or merely certifying them? Are evaluation systems recognizing merit or rewarding pedigree? What AI changes is the plausibility of alternatives. Many intellectual tasks can now be performed outside the arrangements that once made universities indispensable. New systems of publication and review are practical, and independent investigators have access to increasingly sophisticated tools.

The academy need not choose between defending all its traditions and abandoning them. It should preserve the traditions that serve enduring purposes and change arrangements that no longer do. That will require evaluating contributions more thoughtfully, accepting changes in professional roles, supporting different kinds of intellectual excellence, and acknowledging that some formerly scarce abilities are becoming widely available. It will also require scholars, including those of us whose careers are largely behind us, to recognize that the value of our work does not depend on its remaining beyond the capabilities of others. If AI can teach a subject as well as I once did, or help someone make a discovery I could not make, that is progress. Scholarship succeeds when knowledge and understanding advance.

A Renaissance of Scholarship

I do not expect the transition to be easy. Institutions change slowly when existing arrangements support livelihoods, identities, reputations, and long-term work. Legitimate concerns about accuracy, accountability, intellectual independence, and the training of future scholars will complicate the transition. AI may initially magnify low-quality publication, bias, and academic gaming. Some important investigations will remain expensive and institution-dependent. Those difficulties are reasons to govern change carefully, not to assume that existing arrangements must remain unchanged.

Scholarship could become less narrowly associated with academic employment. More people could participate seriously in inquiry throughout their lives, crossing disciplinary and institutional boundaries more easily. Universities could concentrate resources on work that genuinely requires laboratories, archives, field sites, long-term teams, and intensive human mentorship. Students could encounter scholarly work earlier and participate more directly. Some professional positions would disappear or change; other opportunities would emerge; and some forms of support would shift.

Efficiency does not remove institutions’ obligations to the people affected by change. Universities should manage transitions humanely, give employees time and support to adapt, and avoid imposing most of the costs on adjuncts, non-tenured instructors, early-career scholars, and others with the least security. A technology that advances knowledge while worsening avoidable inequities would represent an institutional failure, not an inevitable price of progress.

Those changes should be judged by their consequences for knowledge, education, and society, while recognizing that the burdens of transition are themselves part of those consequences. AI may allow more people to explore, question, discover, and understand. It may restore some of the pleasure of inquiry to a scholarly world burdened by professional productivity. The academy’s achievements deserve respect, but preserving them does not require preserving every practice that grew around them. Scholarship is older than the research university, journals, tenure, citation indexes, and professional scholarly careers. It will outlive many of those arrangements in their present forms.

The arrival of AI should therefore prompt neither the uncritical celebration of academic disruption nor the reflexive defense of academic tradition. Instead, it should prompt a renewal of the question that gives scholarship its purpose: How can we best advance human knowledge, understanding, and intellectual life? If we keep that question at the center, AI need not mark the decline of scholarship. It may help bring about a renaissance.

Author’s note: This essay was developed in collaboration with ChatGPT. The ideas, arguments, judgments, and revisions are Louis A. Bloomfield’s; ChatGPT helped organize them and draft the prose. Bloomfield reviewed and takes responsibility for the final text.

Sources and Further Reading

San Francisco Declaration on Research Assessment. Read the Declaration. Recommendations for assessing research on its own merits rather than through journal-based metrics.

Hattie, John, and H. W. Marsh. 1996. The Relationship Between Research and Teaching: A Meta-Analysis. Review of Educational Research 66, no. 4, 507-542. DOI 10.3102/00346543066004507

Federal Demonstration Partnership. 2020. 2018 Faculty Workload Survey Report of Primary Findings. Full report.

National Institutes of Health. Research Performance Progress Report. Description of progress, impact, changes, products, and public outcome reporting for supported research.

eLife. eLife’s New Model: Introducing Reviewed Preprints. Description of public reviews and assessments attached to citable preprints.

I have become increasingly concerned about cognitive offloading—the tendency to let tools do intellectual work that we once had to do ourselves.

Some cognitive offloading is wonderful. I am perfectly happy to let a calculator do arithmetic, a search engine retrieve an obscure fact, or a GPS remember the turns on an unfamiliar trip. Civilization itself is a gigantic exercise in cognitive offloading. None of us could function if we had to rediscover everything for ourselves. The problem arises when we offload the very activity we are trying to learn.

I encountered this problem long before ChatGPT. For many years, my physics students in How Things Work wrote term papers in which they researched some technological device or subject and explained how it worked. The finished papers were not particularly important. I used to joke that if I had asked the students to turn their papers in directly to a shredding machine, perhaps 95 percent of the educational mission would already have been accomplished. The remaining five percent was their reading of the comments.

The value was in producing the paper: examining the physical objects themselves, finding information, figuring out what was reliable, struggling to understand unfamiliar physics, reconciling contradictory explanations, organizing ideas, and eventually explaining the subject coherently. As more and more information became available online, however, the assignment changed. Students could increasingly assemble papers from material other people had already written. The resulting papers might look perfectly good, but the intellectual activity I wanted was disappearing. I realized that I was no longer teaching physics and scientific thinking. I was teaching text assembly and paraphrasing. So I dropped the assignment.

Generative AI has taken this problem to an entirely new level. A student can now find expert analysis instantly and produce an excellent-looking essay, laboratory report, computer program, or explanation without doing much of the intellectual work that the document or discussion once implied. Instructors can offload their work, too: AI can create assignments, grade them, write comments, and summarize student performance. In the absurd limiting case, an AI writes an assignment, another AI completes it, an AI grades it, and an AI summarizes the feedback. Everyone receives beautifully documented evidence that education occurred. Almost nothing will have happened inside a human brain.

Much of the discussion about AI and education focuses understandably on cheating. It is becoming almost impossible to distinguish student work from that of AI. Another, more optimistic discussion focuses on AI as a teacher or tutor. AI can explain a topic or subject several ways, adjust itself to an individual student, provide endless practice, and respond immediately to questions. Both discussions are important.

But there may be another possibility that is potentially much more disruptive: AI might help save education by making many of our educational proxies less valuable.

Education is filled with proxies. Grades supposedly indicate learning. Degrees supposedly indicate education and ability. Admission to a prestigious university supposedly indicates exceptional promise, though admission itself makes extensive use of proxies. Resumes, honors, recommendations, standardized tests, extracurricular activities, internships, and countless other credentials supposedly tell us something about the person behind them.

Most of these measures began for sensible reasons. But they are subject to Goodhart’s law: when a measure becomes a target, it tends to stop being a good measure. Once everyone wants the grade, people optimize for the grade. Once everyone wants the credential, people optimize for the credential. Students learn what will be on the test. Parents hire tutors and consultants. Resumes are padded. Courses become easier or grades drift upward. Institutions advertise rankings and outcomes. Everyone learns the rules of the game.

No villains are required. A student trying to get into medical school has perfectly sensible reasons to maximize a GPA. Parents naturally want every legitimate advantage for their children. Professors want their students to succeed. Universities want enrollment, retention, happy alumni, and financial stability. Employers need inexpensive ways to sort large applicant pools. The original educational mission can become buried in gameable proxies and individually rational behavior can produce a collectively foolish system.

The underlying problem is that we often measure something because the thing we really care about is too difficult or expensive to measure directly. If what we measure is not really what we care about, then sooner or later we may become very good at producing useless measurements.

But AI could change that.

Imagine applying for a job without initially supplying your college, GPA, degree, or academic honors. Instead, you spend an hour or two with an AI interviewer designed specifically for that job. It presents unfamiliar situations and then follows your reasoning wherever it leads. For example:

• “Here is a machine that has begun failing intermittently. These are the symptoms and these are the measurements we have. What possibilities occur to you?”
• “Why do you think that?”
• “What additional information would you want before deciding?”
• “Here is the information you requested. Does it change your conclusion?”
• “Another engineer thinks the problem has a completely different cause. How would you decide between the two explanations?”
• “You don’t know an important technical detail. How would you go about learning it?”
• “Here is some new information that makes one of your earlier assumptions wrong. What do you do now?”
• “You have ten minutes to become useful on a subject you know almost nothing about. Start asking questions.”

The particular questions would, of course, depend on the job. An AI interviewer could probe knowledge, understanding, judgment, reasoning, intellectual flexibility, communication, persistence, skepticism, estimation, creativity, and the ability to recognize nonsense. More importantly, it could distinguish between what a person already knows and how effectively that person can learn, generalize, and reason when the familiar script runs out.

That distinction matters. I have encountered people who knew an enormous amount about a particular subject but had difficulty moving beyond what they already knew. I have also encountered people who initially knew almost nothing about a subject, but understood new ideas so quickly that they soon caught up with and then surpassed the apparent expert. An assessment concerned only with present knowledge could easily select the wrong person.

Human interviewers can sometimes uncover these differences, but interviewing well is itself a specialized skill. Subject-matter expertise does not automatically make someone a good interviewer. Interviewers can waste time on uninformative questions, become distracted by whether they personally like an applicant, favor people who resemble themselves, or simply fail to probe the characteristics that matter. I know that I am not particularly good at assessing some human qualities in an interview. I can often recognize whether someone is smart, clever, or witty, but other characteristics are much harder for me to judge reliably.

A carefully designed AI system could potentially become much better at the interviewing itself. It could conduct thousands or millions of interviews, compare its predictions with subsequent job performance, learn which questions were informative, and become steadily better at distinguishing knowledge from memorization, confidence from competence, and current expertise from the ability to learn.

The cost matters enormously. It is impractical for a skilled interviewer to spend two hours conducting individualized assessments of a thousand applicants. AI could make individualized examination cheap.

And because AI can generate essentially unlimited new situations and pursue each applicant’s answers in different directions, conventional test preparation becomes less useful. If applicants learn how to answer one sort of question, the system can change the question. If a coaching industry develops techniques for gaming the interview, the AI can study those techniques too.

There will still be a Goodhart arms race. There is probably no permanent escape from Goodhart’s law. But an adaptive assessment can keep moving closer to the thing we actually care about. As the distance between the measurement and the real abilities required by the job shrinks, the harmful effects of preparing for the measurement should shrink too. At the limit, preparing for the assessment increasingly resembles acquiring the knowledge, judgment, reasoning ability, and skills the job actually requires. Gaming begins to look suspiciously like education.

I would suggest conducting two assessments. In the first, put the applicant metaphorically in a Faraday cage—electronic isolation—with no search engine, no AI assistant, and no outside cognitive machinery. What knowledge and intellectual machinery does this person actually possess?

Then open the cage. Give the applicant AI, calculators, search, reference materials, and whatever tools the real job permits. Give the person a much harder problem and see what happens. Some people will simply ask AI for answers. Others will question it, challenge assumptions, notice errors, compare alternatives, perform reality checks, and use the machine to become far more capable than either the human or AI would have been alone.

Now consider what this could do to education. If employers can cheaply measure what applicants actually know, how they think, how quickly they learn, and what they can actually do, the value of gaming educational proxies falls. Getting an A instead of a B matters less if the employer isn’t interested in the transcript. Padding the resume matters less if the resume is barely considered. Attending a prestigious university matters less if the employer asks, in effect, “Never mind who admitted you four years ago. Show me what you can do today.”

I used to joke that one of Harvard’s most valuable features was its admissions office. If Harvard has a reputation for selecting exceptionally capable young people and if almost everyone admitted eventually receives a Harvard degree, employers can reasonably infer that a Harvard graduate is likely to be capable. I called this “proximity to greatness”: once admitted to a respected group, an individual acquires some of the group’s reputation, whether or not the inference is correct in that particular case.

Direct assessment would weaken that inference. Harvard might be wonderful, but tell me about you.

And this could produce a remarkable reversal in educational incentives. If employers stopped rewarding the proxies and started measuring the underlying abilities, students would once again have a strong reason to acquire those abilities. Understanding would matter because understanding would eventually be tested directly. Critical thinking would matter because someone would ask you to think. Historical perspective, quantitative intuition, scientific understanding, judgment, communication, craftsmanship, intellectual independence, and the capacity to learn would become valuable not because they raised a grade but because they made you more capable.

The structure of education might begin to change as well. Four years, 120 credits, a particular sequence of courses, and a particular collection of grades are themselves proxies. Some people might need years of broad education; others might become ready for particular kinds of work much sooner. Students might spend more time on subjects that genuinely enlarge their capabilities and less time accumulating boxes checked toward a credential. Education could become more varied in length and focus because the ultimate question would no longer be “Have you completed the prescribed pedigree?” but “What can you actually do?”

Instructors could then spend less energy policing cheating and managing grading systems and more energy doing what brought many of them into education in the first place: helping students understand things. Universities might even find themselves competing more on how much their students actually learn and less on what I think of as the corporate mission: maximize income, minimize risk, and build the brand.

There are plenty of difficulties. AI assessments could themselves contain biases, measure the wrong qualities, be manipulated, or become detached from actual job performance. Humans would still need to decide what characteristics legitimately matter, provide appropriate safeguards and accommodations, and continually compare assessment results with what happens in the real world. We would create new proxies, and we would eventually begin gaming them. Human nature is not going away.

But perhaps AI gives us an opportunity to move the measurement much closer to the thing being measured. That possibility strikes me as wonderfully ironic.

AI threatens education because it makes cognitive offloading easier than ever before. It can produce the essay without the thinking, the answer without the understanding, and the credential without necessarily revealing what lies behind it. But the same technology may eventually make it cheap to ask directly the question our elaborate system of grades, degrees, pedigrees, and credentials has always been trying indirectly to answer:

What does this individual actually know, understand, and know how to do—and how well can this person learn what comes next?

If we can answer that question reasonably well, much of the incentive to game education may disappear.

And AI, unexpectedly, might help make the purpose of education once again be education.

Acknowledgment

This essay grew from my observations about increasing cognitive offloading and from an extended conversation with ChatGPT about its effects on education. Our discussion began with my concern that students and instructors can now offload activities that are themselves central to learning, and eventually led us to a much more optimistic possibility: AI might reduce the importance of grades, degrees, pedigrees, and other educational proxies by making direct, individualized assessment of human knowledge, reasoning, learning ability, and skill inexpensive and practical.

I supplied the experiences, arguments, concerns, and perspective, including my experiences teaching How Things Work and my longstanding concerns about Goodhart’s law, credentialing, and institutional incentives. ChatGPT challenged some of my generalizations, helped develop the ideas of adaptive AI interviewing, direct assessment, and measuring the ability to learn rather than merely existing knowledge, explored their limitations and unintended consequences, and drafted and worked with me to revise this essay from our discussion.

I have become increasingly interested in a problem that is probably as old as organized human society: how do well-intentioned people, trying to make the world better, end up doing things that are foolish, counterproductive, or even harmful?

I don’t think the answer is that most of them are bad people. Quite the opposite. Bad people are the easy case. The more interesting problem is what happens when decent people—with ordinary human needs for friendship, approval, status, security, meaning, and self-respect—join together in pursuit of ideals.

Humans are social animals. We like belonging. We like being respected. We like feeling that our lives have meaning and that we are on the side of good. A social or political movement can provide all of those things at once: friends, community, purpose, moral elevation, and sometimes professional opportunities, money, influence, or power.

None of that makes the movement wrong. But it creates incentives that can gradually become entangled with the original ideals.

I have come to think of one part of this phenomenon as easy virtue. I use the phrase in my own peculiar sense. Easy virtue is the ability to acquire the feeling—and sometimes the social status—of being virtuous without having to do the much harder work that genuine moral judgment requires.

It can be wonderfully easy to repeat the approved slogan, display the proper symbol, condemn the accepted villain, click “like,” or join the fashionable cause. The reward can be immediate: approval from one’s friends, admiration from one’s community, and the comforting feeling of being one of the good people.

Thinking is harder.

Thinking requires us to ask whether the slogan is actually true, whether the proposed cure will work, what its unintended consequences might be, who will bear its costs, and whether we would apply the same reasoning if the political identities were reversed. It requires us occasionally to discover that people we dislike have a point, or that people we admire are wrong.

That is uncomfortable.

So we often engage in a kind of cognitive outsourcing. We let other people do some of our thinking for us. Most of the time this is essential. None of us can personally verify everything we believe. I don’t reproduce every medical experiment before taking a medicine, inspect every bridge before driving across it, or reconstruct all of modern physics before using a semiconductor.

Civilization depends on trust and intellectual specialization.

But there is an important difference between saying I rely on these people because they know more about this subject than I do and saying I know what to believe because people like me believe it.

The first is a sensible division of cognitive labor. The second is groupthink.

Modern technology has made this ancient human tendency vastly easier to exploit. Social media platforms compete for our attention. Algorithms discover what keeps us watching, clicking, sharing, and returning. Anger, fear, moral outrage, and tribal loyalty turn out to be excellent tools for capturing eyeballs.

No conspiracy is required. A machine told to maximize engagement can discover for itself that an outraged person is often an engaged person.

The result is a powerful feedback loop. We encounter a simplified explanation of the world. People we identify with approve of it. We receive social rewards for repeating it. Algorithms notice our interest and show us more of it. We increasingly encounter people who agree with us, while those who disagree begin to seem ignorant, immoral, or dangerous.

Mass movements and factions are nothing new. Eric Hoffer described much of the psychology of mass movements in The True Believer in 1951, decades before smartphones and TikTok. Thucydides was describing factionalism more than two thousand years earlier. Human nature has not changed very much. What has changed is the machinery available to amplify it.

Edmund Burke worried about another aspect of the same problem: the danger of applying beautiful abstract principles to complicated human societies without sufficient respect for experience, incentives, customs, contradictions, and unintended consequences.

Ideals collide.

Justice can conflict with mercy. Equality can conflict with liberty. Safety can conflict with freedom. Helping one group can impose costs on another. A policy that improves one measure can worsen another. There may simply be no solution in which everyone gets everything we would ideally like them to have.

These difficulties are easy to overlook when an issue has been compressed into a contest between good and evil.

I recently had a conversation with a relative that reminded me how easily this happens. She was concerned about the effects of MAGA politics on schools in Iowa, where she once lived. I didn’t particularly disagree with her concerns, but I pointed out that I had concerns of my own about ideological excesses from the other political direction in schools where our grandchildren live in Somerville, Massachusetts.

Initially, she wasn’t pleased with my comparison. Eventually, however, she acknowledged that I had given her something to think about.

That is about the best outcome I can hope for in a political conversation. Not victory. Food for thought.

I want other people to do the same thing to me.

There is another problem that appears when ideals become institutions. An organization may begin with a straightforward mission: solve a problem. But organizations acquire employees, budgets, buildings, reputations, donors, procedures, and careers. Gradually another mission emerges alongside the first.

I sometimes call this the corporate mission: maximize revenue, minimize risk, and build the brand.

Again, this doesn’t require villains. If I work for an organization, preserving that organization seems completely reasonable. My colleagues depend on it. Our supporters believe in it. And if I sincerely believe our mission is important, then protecting our organization begins to feel morally equivalent to protecting the mission itself.

Mission accomplished can therefore become an awkward outcome.

An advocacy organization founded to solve problem X may discover that solving X threatens the organization’s reason to exist. The problem can then be broadened, redefined, or joined by related or even not-very-related problems. Activists who are particularly talented at organizing may naturally move from one cause to another. Journalists need another story. Politicians need another issue. Universities create offices that acquire constituencies and budgets of their own.

The mission drifts.

This is closely related to Goodhart’s law: when a measure becomes a target, it ceases to be a good measure. Organizations begin by measuring activities because they hope those activities will improve the world. Eventually the measurable activities—money raised, people trained, programs created, policies adopted, followers acquired—can become accomplishments in themselves. They become targets.

James Madison famously observed in Federalist No. 51, “If men were angels, no government would be necessary.” His point applies far beyond government. Institutions must be designed for humans as they actually are, not as we wish they were.

And that includes me.

I have spent much of my life trying to reconcile conflicting ideas. I am bothered by logical inconsistencies and double standards, including my own. When I discover that I am treating an argument differently because I like the person making it—or dislike the person making the opposite argument—I want that to bother me. I also want it to bother me when I haven’t reexamined my ideas or opinions for too long.

It isn’t always pleasant. There is psychological comfort in certainty.

But perhaps the opposite of easy virtue is not cynicism. It is hard virtue: doing the intellectual work necessary to remain uncertain when the evidence is uncertain; listening seriously to people we dislike; acknowledging inconvenient facts; recognizing tradeoffs; and changing our minds when our reasoning fails.

Young people are susceptible to easy virtue because belonging and moral identity are enormously important while they are still forming their views of the world. Older people, however, have no grounds for smugness. We have had more experience, but we have also had more years to become attached to our identities, tribes, resentments, and opinions.

None of us graduates from human nature.

So I don’t think the lesson is “beware of progressives,” “beware of MAGA,” “beware of activists,” or “beware of institutions.”

It is simpler and harder than that.

Beware of certainty that comes too easily.

Ask what evidence would change your mind. Notice when you dismiss an argument because of who made it rather than because of its merits. Be especially suspicious when an idea gives you moral satisfaction, social approval, and a convenient enemy all at once.

And occasionally ask yourself an uncomfortable question:

Am I actually thinking about this, or have I allowed my group to think for me?

That seems like a useful question for any century.

Acknowledgment

This essay grew in part from my reaction to Maya Sulkin’s interview with Sabrina Joy Stevens at The Free Press, available at https://www.thefp.com/p/confessions-of-former-social-justice-warrior. Stevens’s account of her years in progressive activism prompted me to think more broadly about groupthink, institutional incentives, mission drift, moral certainty, and the very human attractions of belonging to a cause.

It also grew from an extended conversation with ChatGPT about those subjects. I supplied the experiences, arguments, concerns, and perspective; ChatGPT challenged some of my generalizations, helped develop several of the ideas, pushed back on my careless or inaccurate statements, and drafted this essay from our discussion.

Grade inflation is rampant at colleges and universities. It’s a natural consequence of several issues:

1. Colleges and universities now treat students as customers, and those customers are unhappy unless they receive what they “pay for” — namely the highest available grades. Given the “customers are always right” mentality active at most colleges and universities, there are negative consequences all around for less than stellar grades. 
2. Assigning serious, meaningful grades is hard work and instructors are not rewarded for that hard work. Instead, they are rewarded for making everyone happy, including students, parents, and administrators.
3. Meaningful grading assumes that student work can be accurately and quantitatively assessed, a notion that isn’t always true. Grading isn’t always a good measure, but giving its importance to pedigrees and credentials, it’s always a good game to be played. 
4. Politics, particularly identity politics, makes assigning meaningful grades dangerous for instructors and institutions. Awarding high grades to everyone is a safe choice.

I’m sure that there are other causes for grade inflation, but those are enough. In this era of grade inflation, I would encourage everyone to ignore transcripts, grade-point-averages, and everything associated with these inflationary measures and judge people on their own demonstrated abilities and merits. Degrees and transcripts are a game and a huge fraction of the population (in the USA) are playing that game. They’re all winning and losing at the same time. It’s wasteful, expensive folly. Instead, ignore degrees and transcripts and have a meaningful conversation with anyone you want to assess. 

In a society that encourages us to view businesses as evil, the greedy adversaries of the people, it’s important to realize that small businesses are themselves barely more than the people who comprise them and they are frequently worthy of our admiration and support, not our hostility and opposition. Small businesses are often the good guys, not the bad guys.

It has become all too easy to divide society into two teams: the people (who are surely good and righteous) and the businesses (who are certainly evil and avaricious); black and white. And in that black and white worldview, a middle ground can’t exist, so non-profit organizations of all sorts (i.e., academic institutions, foundations, and NGOs) demand to be part of Team People. It’s us versus them, everything is high-stakes, and the participants often define who is on which team or the other, even if they have conflicts of interest.

In reality, the stakes are not so high and the distinctions are not so clear or important. It’s a grey world after all. Everyone is human and looking for a path forward that satisfies their own needs, expectations, and desires. We’re not all good guys and bad guys; we’re all imperfect mixtures of good and bad. And one of the most undeservingly vilified categories is that of small businesses.

Small businesses are barely more than the individuals who comprise them. Those businesses share the needs, aspirations, and desires of their individuals, but they have an important measure of official organization demanded by the government. That organization brings with it both positive and negative aspects – a measure of legal separation that enables exploration and risk, and a burden of regulation, expenses, and oversight that encumber such development. It’s a mixed bag and no one should underestimate the complexity of trying to launch or maintain a small business.

To dehumanize small businesses into entities that we should treat as adversaries is just plain wrong. While it is convenient and even desirable for modern society to treat all businesses as sterile, never-touched-by-human-hands entities that sell products or services in hermetically sealed and perfectly uniform packages, that is a silly and counterproductive conceit. Nearly everything is touched and affected by humans and that is a good thing, not a failing. Small businesses involve real people and many or even most of them care that you are happy, satisfied, and supported by what they provide. Don’t expect inhuman/antihuman treatment; instead, expect human treatment and respond positively to it when it is good and makes the world a better place.

So where does “no consequences” fit into this essay? It is in the misassignment of what is human and what is less so. Treating small businesses as though they are large businesses, with nameless hordes of people who can be treated as objects rather than human beings, is a serious mistake. Even some large businesses are more human than we typically admit. People are people, and the closer those people are to whatever products, services, or other activities they provide, the more they deserve and value being treated as people. As the co-owner of a tiny business (Adaptive Rubber LLC and its products at EarJellies.com), I can say from experience that working with people as individuals and sharing the human experience with them is more valuable than anything else on either end of any transaction.

Furthermore, expecting grey-area organizations to be “good guys” because of their titles, names, or categories (e.g., academic institutions, non-profits, NGOs) is uncertain at best. Organizations tend to drift away from their original missions once they become large and their workforces begin to think of themselves as employees. Businesses, companies, and organizations become problematic when their employees begin to lose sight of their missions and begin working primarily for incomes rather than purpose. It’s okay for people to earn a living; they need it and it wouldn’t make sense if they did otherwise. But it’s less okay for people to work only for incomes and say to hell with the people at the other end of any transaction or interaction. For the most part, small businesses keep track of the former behavior and deserve our praise and support.

Grades used to mean something in colleges and high-schools, but not anymore. In most schools and most disciplines, human nature has done away with meaningful grades.

Most college (and high school) classes are graded. In the past, those grades served two important purposes. First, they motivated students to work harder and put in the time, energy, and effort necessary to acquire a real education. Second, they provided some lasting measure of the students’ performances: how much they’d learned, how well they thought, how deeply they engaged, how motivated they were, and everything people used to mean by the term “education.”

But meaningful grading is hard work and, in the absence of strong incentives, it falls by the wayside. Meaningful grading means that some students won’t get A’s, or even B’s, and some of them might not even pass the class. People are different in so many ways and their performances in educational contexts are inevitably different. Meaningful grades are intended to encourage students to do their best at the time and then to reflect the differences in what they did.

Alas, meaningful grading has some uncomfortable consequences—ay, there’s the rub¹. Current students believe that education is all about getting high grades and are unhappy (or even outraged) when they receive anything less. Faced with a B (or less), they complain and browbeat their instructors for regrades, do-overs, extra credit and other special deals that require more work by the instructors. Their parents complain on their behalf and go over instructors’ heads to get what they want. And other institutional and societal constituencies apply their own pressures in advocating for the students they support. There is lots of academic privilege and entitlement around these days.

Thus, instructors who spend the time necessary to give meaningful grades must then deal with pushback and that grief consumes their time and peace of mind. Moreover, instructors are typically reviewed and rewarded based in part on student feedback, so there are direct negative consequences (i.e., punishment) for making students unhappy. Since there are no consequences (and less work) to giving out As and substantial costs to anything less, it’s no wonder most instructors don’t sweat over grades anymore. Let them all have As! ²

Lastly, institutions (I’m thinking primarily of colleges, but also some high schools) are rated in part on graduation rates and how quickly students earn their degrees. Those ratings are important to the institutions, which like all companies and many organizations want happy customers and seek to attract more. Since there are no consequences to awarding degrees to less than educated students, why worry about something as inconsequential as the integrity of grading? We’re “just trying to keep my customers satisfied, satisfied.”³

It should be clear at this point that meaningful grading is an ordeal for everyone. Students don’t like it because they’re in school to have a pleasant four years (I’m thinking of college), not have to work too hard, get the highest possible grades, and receive a degree—that golden credential that will lead inevitably to success in life!? Instructors don’t like meaningful grading because it requires so much additional work and the potential of punishment. And institutions don’t like it because it upsets the smooth flow of students through the years to graduation and upsets people in ways that threaten income and ratings, and introduces various types of risk.

So grades inflate and everyone is happy! It’s a no-consequences celebration! Students are happy because they get As and they will surely be rich and famous going forward. Parents are happy because their kids are obviously doing so well and will be wildly successful. Instructors are happy because they don’t have to work as hard assigning meaningful grades to student output that is increasingly juvenile now that high grades are so easily obtained. And institutions are happy because all their students are excelling, paying tuition, graduating promptly, and touting their schools as great and fun and “the best years of their lives.”

Given the various incentives and disincentives, it’s no wonder that grades have inflated into a meaningless mush here in the United States over the past 30 or 40 years. It’s win-win-win!

Okay, there are a few minor consequences, notably a changing in the amount of actual education delivered per (constant) dollar spent. But I’ll focus only on difficulties meaningless grades impose on the job market. For decades, grades and degrees were used for important purposes, notably identifying people who thought well, knew and understood much, and had strong motivation and character. They were also used to discriminate against certain people without violating laws. But now that everyone looks wonderful on paper and college degrees are abundant, what can a potential employer look for when deciding whom to hire?

The answer is straightforward but requires effort: employers must set aside meaningless grades and degrees, and investigate what is going on between each candidate’s ears independently. They must do their due diligence and study what each candidate knows, what they understand, how they think, what is their motivation and character, and what they will bring to the job that is before them.

For jobs that don’t actually require college educations, employers should stop requiring college degrees—why prefer people with meaningless credentials to others who are willing and able to do the job well and possibly better and have shown the good taste not to waste time on a worthless degree?

For jobs that really do require a college education, assess that education! Ignore the (meaningless) grades and instead ask the candidates to demonstrate their knowledge, understanding, thinking, character, motivation, passion, morals, ethics, and intelligence. Don’t be blinded by the flashiness of their credentials—elite degrees have become hollow luxury goods, equivalent to designer bags of dubious quality—and do a serious assessment of the person behind the façade. These days, no one should judge people by their meaningless educational credentials any more than by the color of their skin. Instead, judge them “by the content of their character.”⁴

1. (thanks, WS and Hamlet)
2. Marie Antonette
3. Simon & Garfunkle
4. Martin Luther King, Jr.

Why are there no consequences that…

nearly all non-profit organizations eventually adopt corporate missions.

Non-profit organizations, including charities, professional societies, and educational institutions, are usually founded by idealistic people who want to make the world a better place and who truly believe that their organizations can help to achieve that goal. If asked, those founders can probably articulate a clear mission for their organizations and they would probably not identify making an income as the primary focus those missions.

But once an organization begins to hire a permanent paid staff, the original mission inevitably comes under pressure. When push comes to shove, the staff is invariably more concerned about being paid than it is in pursuing the founders’ vision for the organization. Contributions that were initially directly almost entirely to the organization’s original mission are redirected more and more to meeting payroll and the organization’s mission evolves into a corporate mission: maximize income, minimize risk, and build brand. For the paid staff, receiving an income is the mission and the original mission becomes vestigial, a historic artifact that is used to encourage continued donations or to justify the organization’s non-profit status.

While there are a few non-profit organizations that seem to weather this sad consequence of human nature, I have been amazed at the variety of organizations that have succumbed to it. Nearly every organization that contacts me for money or in-kind contribution is doing so to meet payroll first and to pursue its original mission second (or third, or not at all). Its agents employ the mass-marketing tools of modern technology while exploiting the politeness hammered into many of us during an earlier, more personal era (“Hello. How are you doing today? I’m calling on behalf of…”). Calls from paid solicitors are particularly dispiriting because they can’t even pretend to care about the original mission.

Over the ages, wise people have commented on the philosophical value of separating the person who gives charity from the person who receives charity. Charitable organizations, and non-profits in general, were intended to provide that separation. It’s sad that human nature makes it so difficult for these organizations to avoid losing sight of their original purpose.

Lou Bloomfield

Why are there no consequences that…

universities consider time and money spent in the classroom as wasted.

Universities are much more inclined to “invest” in beautiful facilities and grounds, institutes and centers, sports and entertainment, and information technology. That’s probably because every decision at a university is a financial decision and attracting donations is a central mission of every university. Donors want to see tangible results from their donations; they want to buy something lasting. Ultimately, they’re looking for a little immortality, hopefully with their names attached. Since money spent on teaching is gone at the end of each semester, few donors care to support it or even pay much attention to it. And if it doesn’t attract donors, universities aren’t interested in spending their money on it, either.

Students are pragmatic and they recognize that society values a college degree far more than it values a college education. Students therefore pay for a degree, not for an education, so the quality of the education at a university is essentially irrelevant to all the stakeholders. If education happens, everyone eagerly claims credit, but it’s the result of volunteerism and youthful idealism. Universities and their administrators much prefer to spend their precious budgets on activities that promise to reward the universities and their administrators. Education is not such an activity. Faculty similarly have little incentive to spend time on teaching because they are rewarded for research and scholarship. Lastly, many students are happy to obtain a degree without putting in much effort or obtaining much of an education. Strange world.

Lou

This is a work in progress…