In May of 2021, mid-pandemic, I published a post on my personal blog called “Why Grades are Bullshit.” I wrote it after another day of entering points into a gradebook I did not believe in. That site is gone now, but the Wayback Machine let me go back and see it. A study published this month by some assessment researchers sent me digging for it. Their finding: institutions are redesigning everything about assessment for the age of AI except the thing students, parents, and employers actually see. The grade.
I am republishing the 2021 post here. Then I want to talk about what has happened since, because we have been handed a second once-in-a-generation chance to fix how we represent learning, and we are blowing it… again.
May 2021: Why Grades are Bullshit
I hate things that are pointless. I mean, what’s the purpose of these subjective numbers we use to assign value to a task? They do not mean anything. Letter grades do not even really mean anything. A student who is going to do well in my class gets an A, a student who just wants a credit gets a D. There isn’t really too much in between. So I go through these motions just so we can report the level of effort back to a district who needs a number/letter they understand for a transcript.
Ignoring the subjective nature of grades in general…
Let’s say I take 2–3 points off an assignment for formatting, typos, grammar, spelling, whatever; I still take the time to write to the student why I took those points off. In most cases, they already know they did it right? They rush, they get sloppy, it does not matter to them, whatever the reason… We have to let them know why they lost points and even though they know why, if we do not tell them, they ask (even though if they looked over their submission they would already know). Some hope I will miss it, or look over it; when they get called on it some will ask to fix the issue for those extra 2 points, but it was worth the attempt to get away with less work. I get that. My point is, they will already have my written feedback on what they did wrong. For those who want to do the bare minimum for a credit, they do not care about the feedback they just want to be done and put it behind them, other than to know that they are performing at a minimum level to get by. It is a waste of my time, and theirs. Grades enable and encourage that game.
I also know that those games people play to get credit (even Pass/Fail) is rigged toward the privileged who know how to play; and they will be just fine regardless of grade. Their parents knew how to play it and none of it is reflective of human connection, socialization, insight, or academics. The credit, grade, time are just hurdles or tasks that we have people jump over that keep marginalized communities down.
Even Standards Based Grading, which I think gets us closer to something real that measures ability is still a problem, we still assign a level for the purposes of reporting credit, the feedback is just more relevant and detailed in those reporting periods. In my two decades plus of teaching students who struggle, those with disabilities, those from marginalized communities, I have met very few that would not ultimately be capable of doing any of the things I ask well. But we’re driven by Carnegie units, credits, and grades. For those who truly cannot, their pathways for career and society are already largely set. They get Pass/Fail options, modified graduation requirements, social promotion, etc. For those who know how to form the relationships needed to get ahead, if they have the motivation/privilege, then they will land in their own roles. There are countless examples of straight, white males who continue to fail upwards and gain promotions and raises throughout their career because that’s just what happens in corporate America for those in the dominant culture. The research on people moving between socio/economic barriers doesn’t lead me to believe that very many can change these predetermined societal roles, or economic status.
Grades are part of the fiction, the mass delusion we use to keep some semblance of order. Structures and systems comfort the masses. This is part of why COVID-19 shutdowns of schools freaked out so many to the point where they demand that schools reopen as normal even though the research coming out about reopenings is showing us that schools in fact are mass spreaders as well and that students are efficient transmitters of the virus. When the narrative is disrupted we freak out, we have anti-maskers, people gathering in spite of what science tells us. Demands to go back to the system from which they benefited. These same vocal activists had little to say when graduation rates in Minnesota for students of color was the worst in the nation; or when biased testing, and lack of access to honors/AP courses was the norm. Schools even handed out credits for minimal amount of effort when it was white suburban kids who started to struggle with how to play the new game that would actually have them demonstrate mastery of a skill. We could have used the time to connect people, to teach, to create experiences in new ways (that by the way, where happening for decades in true online learning programs), but we focused on all of the wrong things and not what COVID-19 taught us about what was bullshit about the whole system and how inherent the inequities are, and will become further cemented as we transition back to “normal”.
So, even though I don’t think grades mean anything, I give them to dozens of kids a day because in the end, we are providing a service to the districts the students come from as well as the students we serve. It’s part of our contract with society, we keep learners in their predetermined tracks and roles that our white supremacist, misogynistic, homophobic system demands of us. Disrupting that system is impossible without a competing narrative people can live in and share. More and more are waking up to just how pervasive the inequities are, but is it enough to change our ways? I suspect not. When are we going to be disrupted to this degree again and have the chance for change? It’s the slow gradual incrementalism that wears on me. We have the knowledge to do better and we don’t.
I feel like a sellout every time I have to enter points into a gradebook. What I value about teaching is the back and forth, the experience, the connections, the conversations, the improvements learners make throughout their time with me, and I see enough of it to keep me going. I just try to put all that stuff above out of my head because it’s part of the mass narrative about how we measure kids. And that form of measurement is destructive to individuals, communities, and society.
The week I wrote that
That post did not come out of nowhere. The same week I published a companion piece called “What’s Next for Education?” I was watching districts respond to the shutdowns by grabbing whatever free tools vendors dangled instead of investing in teachers and instructional design, and I wrote: “This was our chance to change it all, not run to Google, EdPuzzle or other vendors for solutions.”
I closed that post with a prediction: “The next big phase of public education is coming soon, and we are not ready for that either.”
Eighteen months later, ChatGPT launched.
Pass them if they try half
There is a part of the pandemic story I have never written about. During the shutdowns, I was directed to give students passing grades and credit if they attempted at least half of an assignment. Attempted. Not completed, not demonstrated. Attempted.
I understood the reasoning and I complied with it. Students were in crisis, and punishing them with failing grades during a global emergency would have been cruel. But sit with what that directive admits about the sytem. If a passing grade can mean “demonstrated the skill” one year and “opened the document during a catastrophe” the next, then grades were policy levers all along, adjustable whenever the adults decide compassion outranks measurement. The institution said so out loud. Would it just expected everyone to forget once the buildings reopened? Yes.
And it was not only triage. In the summer of 2020 I wrote a post called “Differences Between Distance Learning and Online Learning,” comparing the emergency scramble to what established online programs were doing. Here is what I documented at the time:
“Many online programs changed grading practices and requirements for credit attainment early on to support the needs of their learners. Demonstration of skills aligned to academic standards was still necessary, but there was no longer any point scale or letter grade acting as a barrier. Teachers started counseling and cheerleading effort (this results in greater levels of participation and engagement).”
We removed the point scales and letter grades, kept the requirement that students demonstrate skills, and participation went up. That was a running experiment with results, documented while it was happening. After we could all see the results, we put the barrier back.
What was true about students then that is not true now?
The disengagement did not end when the masks came off. Anyone teaching right now can tell you the anxiety is worse. Students are carrying academic gaps we never remediated, a mental health crisis we never staffed for, phones engineered to hold their attention better than we do, political and economic instability at home, and now a technology that will happily complete the assignment for them while they feel nothing. Every condition that justified the mercy of 2020 still holds. The mercy is what ended. No one can possibly believe that we took it back because students recovered. We took it back because the system demands its sorting mechanism again.
The elephant is still in the room
Which brings me to the study that sent me into my own archive. Seven researchers looked at how higher education is responding to generative AI and asked the question: has anyone rethought grades?
They reviewed the publicly available policies of 55 public universities across Australia, Finland, and Hong Kong. Nearly every institution had guidance about AI in assessment. Almost none had touched how student achievement gets represented. In their words, there is “a blistering silence regarding reimagining grades,” while “entire programs of assessment are being reworked with little to no thought given to grades.”
They also name the pattern I lived through. COVID pushed institutions into softer emergency grading, and then, as they put it, “grading returned ‘back to normal’.” The sector ran its dress rehearsal for rethinking what grades mean, under the most forgiving conditions imaginable, and filed the results in a drawer.
Dead ends
The study sorts the current responses into two camps, and if you work in a school or district you will recognize both, because both are being sold to you right now.
AI-proof grading is the lockdown response: proctoring software, detection tools, handwritten exams, oral defenses, anything that guarantees students did the work without AI. Notice what it protects. As the authors put it, the purpose of AI-proofing “is precisely to safeguard the existing values, norms, and practices of grades.” The fortress is being built around the gradebook.
I happen to believe that AI-proof anything is next to impossible when it comes to assessments over the course of a class.
AI-powered grading is the efficiency response: let the machine score the work, faster and supposedly free of human bias. Vendors pitch it as liberation from the grading burden. Automating a measurement no one believed in just makes the measurement cheaper and faster.
Both camps answer the question of how to keep grading in the age of AI. Neither asks whether the grade was worth keeping.
It’s worse in K-12
The study is about higher education, but everything in it is harder in K-12, where I spent my career. Universities at least have regulators nudging them toward whole-program assessment redesign. We have Carnegie units. Seat time. Credit accrual rules written before the internet. Transcript formats no district can change unilaterally, because colleges, athletic associations, and state systems all consume them.
And our version of the two dead ends is already funded. Districts are signing integrity-tool contracts, which is AI-proofing, and piloting AI grading assistants inside the LMS, which is AI-powered grading, while the report card that comes out the other end remains what it was a generation ago. Standards-based grading, where it has been adopted, gets closer to something real. I said so even in my 2021 rant. It still ends in a reported symbol, because the machinery downstream demands one.
Meanwhile the students most harmed by the current system are the same ones I wrote about five years ago: the ones for whom grades track privilege, compliance, and knowing how to play the game long before they track learning.
Five questions for your district
The researchers close with five reflective questions for institutions. They wrote them for university policymakers. Here is what they sound like translated for a school board, a superintendent’s cabinet, or an online program:
Who actually consumes your transcripts and report cards, and what decisions do they make with them? If the answer is “colleges and compliance,” then who is representing what students can actually do?
What achievement is worth putting on the record? Your students do real things that never appear in a gradebook, and your gradebook is full of numbers that reflect deadlines and home internet access as much as learning. Which one is the record?
What can anyone do with the data you report? A “B” travels easily between institutions precisely because it says so little.
Who gets to build the record? Right now it is teachers, alone, under time pressure, which is exactly why someone is trying to sell your district an AI grader. What would it mean for students to help construct the representation of their own learning?
What can your students do with AI, and without it, and does anything you report tell the difference? A transcript cannot currently distinguish the student who directed AI skillfully from the one who copied and pasted, or from the one who never touched it.
Power concedes nothing
I ended the 2021 post by calling myself a sellout for entering points into a gradebook. I ended its companion by admitting I was burnt out. I left the profession because I could no longer wait for administrative support around the challenge posed from grading AI submissions and the system is never going to change itself.
In that companion post I leaned on Frederick Douglass: “Power concedes nothing without demand.” Nobody with authority over transcripts is going to volunteer to rethink them. Grades will survive AI exactly the way they survived COVID, by default, unless the people in the rooms where assessment policy gets written put the question on the agenda. You do not have to solve it this year. But can we refuse to pretend it is settled?
We wasted the last disruption. This one is bigger, and it is not going to wait for us either. Let’s not wait for the third once-in-a-generation chance to make change.
If your district or organization is rewriting assessment or AI policy and wants a thinking partner who has lived both sides of the gradebook, book a consultation with Inclusive AI Strategies or subscribe to the newsletter.
Reference
Nieminen, J. H., Bearman, M., Boud, D., Corbin, T., Dawson, P., Tai, J., & Walton, J. (2026). Grading in an age of assessment reform: the elephant in the room. Assessment & Evaluation in Higher Education, 1–17. https://doi.org/10.1080/02602938.2026.2694451
