I just finished reading Paul LeBlanc’s book Students First: Equity, Access, and Opportunity in Higher Education.
To be clear, this post is neither a book review nor a comprehensive exploration of Competency-Based Education (CBE), although it contains some elements of both. Rather, I’m interested in exploring how to think about a well-drawn and incredibly ambitious proposal for change in education. How can we think about change in such a complex and—up until very recently—stable system? In particular, Paul refers briefly to the notion of an “ecosystem.” I’m going to take that analogy seriously. Nature has powerful mechanisms for both ensuring stability and driving change. They are amoral in the sense that the only outcomes they drive toward are the survival of species and resilience of ecosystems. There is no “good” or “bad.” Nevertheless, they offer powerful mechanisms that are echoed within our human systems, often without us fully realizing it.
How to read this book
Before you read this book, I recommend that you wait until September 27th when Paul’s new book, Broken: How Our Social Systems are Failing Us and How We Can Fix Them, will be available to buy and read that one first. I was privileged to receive an early copy. It is one of the most important and morally courageous books on education that I’ve read in years. (I’ll have more to say about it after its official publication date.) It will also make Students First more legible.
This is a trick I learned from Paul himself. He was recently generous enough to make time for me to seek his advice on a matter of some importance to me. Before we started, he took the first 45 minutes of the conversation to coax my life story out of me. While he was genuine and warm, he was also practiced at it. I soon found out why. He proceeded to use everything he had learned about me—my values, my ambitions, my way of thinking, my emotional compass, and more—to help him understand what I was trying to do and why I was trying to do it. He was uncannily effective at it.
CBE is a fraught topic for a lot of reasons. It’s hard not to bring some baggage to the conversation. Broken is a deeply personal, at times biographical, and occasionally confessional book. By reading it, you will understand Paul better. That, in turn, will help you understand Students First better. Paul is a broad thinker and, given the complexity of the topic of CBE, his style of taking on all arguments and covering all cases can easily be misread, particularly if you are already ambivalent about the subject (or actively opposed to it). If you read Broken first, you may still disagree with Students First, but you will read it more sympathetically because you will understand the values and the commitments to education of the person writing it.
These two books are well worth your time. They’re short and well-written (though they reward more careful reading and thought than their easy-to-read prose might lead you to believe). I encourage you to read them both.
The nub of CBE
At any rate, on to the topic at hand. At a high level, CBE poses two questions:
- Why should we design a system that prioritizes how long students spend in class over how well they learn? In other words, if one student comes in already knowing half the material in the class and another comes in needing a little extra time to master it, shouldn’t we design our education system around accommodating both of these students?
- Why should we prioritize content coverage over skills learned? In other words, isn’t it more valuable for a student to be able to analyze an unfamiliar work—a work of literature, an economic argument, a biological experiment, a musical composition, etc.—than it is for them to be familiar with a list of such works that somebody compiled for them?
Time versus mastery and knowledge versus skills. That’s how I understand the nub of CBE. I’m sure the CBE experts reading this will find flaws in this reductive formulation—please feel free to do so in the comments thread—but as a first approximation, these are the challenges that CBE poses to the traditional educational model. I find it hard to argue with these principles. The devil, as usual, is in the details. But those details are not the primary concern of this post.
I’m going to explore the mechanics of CBE in a little more detail later in this piece. Just a little. First, though, I want to frame up the notion of an “ecosystem,” which is more central to my thinking as I write this. As a thought experiment, please accept for the sake of argument that CBE is a positive innovation and worth propagating.
How would we do that? How do we even go about thinking about how to do that?
Ecosystems
Paul raises the topic of ecosystems early with an explicit shoutout to evolutionary biology:
To borrow from E. O. Wilson, the renowned conservation biologist, “insular biogeography” is the idea that isolated ecological areas, being disconnected from each other, are inherently less robust and supportive of life than healthy ecosystems. Our institutions of higher education largely exist as islands and thus are not serving students who try to navigate their educational journey in a world that promises and rewards mobility.
Students First: Equity, Access, and Opportunity in Higher Education
When Paul refers to “islands,” think “Galapagos Islands.” Each university is its own precious, isolated ecosystem that is vulnerable to external shocks. In nature, it might be an invasive species coming to an island. In education, it might be a change in demographics or a pandemic that has long-term effects on student enrollment patterns and modality preferences.
From here, Paul immediately launches directly into a discussion of data interoperability. How do we connect university ecosystems? Through common agreements about student achievements and technical interoperability standards that support sharing of data about student achievements.
He’s not wrong. That said, ecosystems are generally not architected. They evolve. Sometimes they can be cultivated. And they always follow certain rules that are worth exploring because they explain both change and stability in complex contexts like education.
For starters, ecosystems in nature are driven by one imperative: natural selection. Survival. Individual critters within the ecosystem strive to survive, reproduce, and have their offspring survive and reproduce. Sometimes they randomly mutate. The mutations may enhance their chances of survival, hurt them, or be neutral. In my backyard, many species of sparrow co-exist. Their variations provide them with neither an advantage nor a disadvantage. As such, natural selection doesn’t kick in. It doesn’t care. It only exerts its influence when competing variants become either more or less well suited to survive and reproduce within a particular ecosystem.
Survival does not exist in a vacuum, thanks to predator/prey relationships, competition for resources, symbiotic relationships, and other interactions among species within an ecosystem. In fact, the sum of these interactions is the ecosystem. And it has within it feedback mechanisms. For example, when field mice thrive and multiply, so do predators who feed on them such as hawks and foxes. Conversely, when the mouse population drops, so does the predator population. For those of you familiar with systems thinking more generally, ecosystems function through reinforcing and balancing feedback loops. Some actions are amplified while others are balanced out.
These feedback mechanisms give ecosystems resilience. Paleontologists Niles Eldredge and Stephen Jay Gould proposed that ecosystems exist in a state of punctuated equilibrium. Normally, the feedback mechanisms keep the ecosystem in a state of balance. Much like a thermostat, the balancing feedback loops help the system adjust to deviations. While the system does fluctuate, it stays within an acceptable range. But external shocks to the system like an invasive species, human impact on habitat or a keystone species, or climate change can cause sudden, dramatic, even violent shifts in what may have seemed like a permanently stable system. The changes driven by reinforcing feedback loops will continue until the ecosystem finds a new balance in which novel feedback loops among the surviving species achieve stability (or until the ecosystem dies completely). A virus will continue spreading, multiplying, and mutating until its hosts develop strong and broad enough immunity to keep it in check. We all have experienced the havok such a dramatic shift can wreak in a short period of time.
In evolutionary biologist Richard Dawkins’ seminal work The Selfish Gene, he proposed the idea that the unit of life that is competing to survive is not the species but the gene. For example, the genes on the spike protein of the SARS CoV-2 Omicron variants outcompeted the genes of the original strain and thus became the dominant strain.
When we use “ecosystem” as an analogy in education, it’s sometimes hard to figure out which entities are the ecosystems, which are the species, and which are the genes. In Paul’s usage, each institution is its own ecosystem. As humans, we tend to focus most naturally on the species level. We recognize different types of plants and animals. It’s hard for us to hold in our heads the level above—i.e., the feedback loops among multiple species and the physical environment that make up an an ecosystem—or the level below—i.e., genes that express themselves in complex and non-obvious ways to influence the species.
This may seem like a lot of detail but it will come in handy when we think about a theory of change for a shift in punctuated equilibrium like CBE.
Let’s put a pin in this idea and return to it shortly after a very brief detour.
CBE, backward design, and natural selection
Another touchpoint that Paul mentions briefly but is worth dwelling on is backward design. The basic notion is as follows:
- When designing your curriculum, first think hard about meaningful and observable learning outcomes. What do you want to make sure your students achieve? What would it look like for them to achieve it?
- Next, drill into that second question and design authentic assessments. How can we give students meaningful, real-world ways to demonstrate that they’ve learned what we’re trying to help them learn? How can we make sure our assessments are truly meaningful and reliable?
- If you’ve nailed the first two parts, then you can be pretty wildly experimental with how you teach. You’re always measuring against results using the same yardstick.
Paul returns to this last point repeatedly. Well-executed CBE is entirely agnostic about teaching methods. Whatever teaching method helps students achieve competency is good by definition.
(By the way, my characterization of backward design may be as crude as my one of CBE. Corrections are welcome in the comments thread.)
Notice the analogy to natural selection. In a balanced ecosystem, a course design mutation that results in poor student results against the competencies should not survive. A design that produces better results than others should outcompete the others. And ones that achieve similar outcomes should be able to coexist.
Keep in mind, too, that we can have multiple ecosystems and even micro-ecosystems. A claim of “achieving similar results” begs the questions “for whom” and “under what conditions”. Even one ecosystem can have a stream, a meadow, a forest, a vernal pool, and so on. It’s not one-size-fits-all. Each micro-ecosystem within the larger whole has its own conditions for survival. Each is well-suited to certain kinds of species with certain genes moreso than others. Likewise, different students will thrive in different educational ecosystems.
Paul’s rhetorical pivot from ecosystems to technical interoperability comes from E. O. Wilson’s point about interactions across ecosystem boundaries. A frog starts its life as an aquatic animal and later comes to live on land. You might think of it as a transfer student. It needs to bring everything it developed in the aquatic environment (like its legs, for example) into its next environment to thrive. Birds and other animals migrate, sometimes thousands of miles, to spend different parts of their lifecycles in very different environments.
Both the ecosystems themselves and the creatures that pass from one to another are more resilient when this kind of transition works. Students moving from a two-year college to a four-year college are more likely to complete their degree if all their credits transfer. This makes for more successful students and more successful universities. Likewise, students who graduate from college to work effectively are more likely to succeed if the skills they learned at college transfer to the work world, to the mutual benefit of the graduates and their employer. Every time a person passes successfully from one ecosystem to the next, that person brings resilience and diversity of thought, skills, and lived experience that strengthens the new ecosystem. On the other hand, closed ecosystems where species never move to or through the boundaries tend to result in species—and the balance of feedback loops that constitutes the ecosystem—that are more vulnerable to external disruption. The less that colleges are enmeshed through transfer agreements and the less often they have students that are highly successful at crossing boundaries with adjacent ecosystems (like K12 and the workplace), the more fragile they will be. In our cultivated human ecosystems, technology can help (or hinder) the connections among and pathways between the environments we create.
What’s missing
Paul runs through a wide array of of successful CBE examples. He writes about a seminary school that uses CBE and argues—compellingly, in my view—that philosophy can be taught with depth and integrity via CBE. In fact, speaking as a philosophy major, if my program had focused more explicitly on how reading Plato and Aristotle, Locke and Hume would teach students how to read closely, think critically, write clearly, argue compellingly, and change their minds based on evidence and argument (i.e., learn), the department likely would have attracted more philosophy majors rather than fewer. I can say with great confidence that my training as a philosopher gave me competencies that have served me well in a variety of jobs that did not yet exist at the time that I was in college.
I got lucky. Nothing in my college drove me toward philosophy or showed me how it could make me successful and resilient. I was not meaningfully tested in ways that made me more likely to gravitate toward skills that would be useful to me throughout my life. From an ecosystem standpoint, we are thin on single natural selection drivers. In the book, Paul does identify drivers toward competencies in certain critical disciplines, noting that one would not want to fly with a pilot who passed every test except landing or a surgeon who is competent in everything except identifying organs. A second area he talks about is credit transfer, where institutions are rightly under increasing pressure to fix a broken system that forces students to redo work, costing them time and money, because their courses from their old institution are not accepted by their new one. And a third area is stackable graduate micro-credentials, where working professionals want to build toward graduate degrees in small, manageable, and useful increments. These, along with the hypergrowth of pro-CBE institutions like SNHU and WGU, are areas where demonstration of competencies can serve as the driving measure of success that helps create a reward system for innovation. But in places where there isn’t a “competency or die” mechanism, there will be no evolution toward competencies and no variation and natural selection that will help us to discover what thriving CBE-driven ecosystems can achieve for the humans that inhabit them.
This is the top-down part of evolution. The forces that shape survival and reproduction. Up until very recently, there haven’t been strong pressures on higher education to change in general. To adapt to its environment. To adapt, in particular, to the needs of its students. To the contrary, the modern university evolved to achieve stability against political pressures for change. The multiversity—the model for the modern university that was conceived of as part of California’s Master Plan—is a resilient ecosystem that maintains stability in part because it was designed to avoid certain kinds of competition within the ecosystem. Master Plan architect Clark Kerr did not have an economic problem when he created the University of California model that now dominates the world of university designs and rankings. The California state legislature was both able to and inclined to invest. No, Kerr’s problem was political. He faced factions, both within the university and within the state legislature, that felt in competition with each other. Even though there was money for everyone at the time, competition for status will always be a zero-sum game. So Kerr invented the multiversity to discourage comparisons that might lead to competition. Universities, he argued, could serve their research, teaching, and community mission goals. Just don’t look too closely at the accounting. If you examine who is getting how much or hold one group too visibly accountable for performance, that creates strife that threatens the coalition holding the modern university together. In a way, today’s universities are not institutions at all and are barely ecosystems. Each department and function is its own Galapagos Island. Which is why it can be maddenly difficult to try to get traditional universities to do anything at an institutional level. The “institutions” are barely more than fictions. They are not in a state of balance so much as they are in one of stasis, largely powerless to change in the face of either internal or external pressures.
The multiversity and the status hierarchy it creates among universities is a strong environmental condition that has made the ecosystem of higher education as a sector incredibly stable. It, along with the Carnegie Unit and the Higher Education Act of 1965 cited in Students First, create incredibly strong balancing feedback loops to add stability while dampening reinforcing feedback loops that could drive change. The multiversity inspired a series of ranking systems and, more broadly, a status system that ensures more money and status flow to R1 universities. It creates a clear survival imperative along with rules for thriving in the ecosystem. Don’t want to starve? Then don’t neglect research. While I am no policy expert, I do not see any strong survival incentive in this ecosystem for measuring and improving student outcomes (which is likely why Paul, who is far more of a policy expert than I am, spends siginficant time on policy proposals in Students First). Some pressure has come increasingly in recent years in policies that focus on student graduation rates and time-to-completion, but those are not really measures of accountability for learning. In fact, time-to-degree is partly driven by state budget concerns in a public education system where every student’s tuition is subsidized by taxpayers.
On a more micro level, the humans who want to survive and thrive in this system are driven by the imperatives they are given by the ecosystem. Unlike genes, they are conscious. Their changes are not random. They have agency and intelligence. But at the end of the day, the people living and working in a system cannot ignore needs for economic security, status and approval, a sense of professional identity that makes them feel good about themselves, and so on. If we want an ecosystem that drives toward improving student outcomes (whether through CBE or some other means), then we need to change the environmental conditions such that educational innovation and measurement of success at some clear measure of progress are rewarded. Currently within higher education, the humans working in the system often lack the natural incentives in their teaching that both rewards effective innovations and provides vectors for those innovations to propagate.
Contrary to popular belief, research supports the position that even tenured, research-oriented professors tend to care deeply about the quality of their teaching. But they are have neither the motivation nor the tools to improve beyond certain bounds. From a motivation perspective, they fear “experimenting” on their students (which I extrapolate from the research to mean deviating from the teaching methods that they are comfortable with and believe have been effective for them). They also fear looking like idiots in front of their students. And if they wanted to experiment, how would they know that they are succeeding? What tools do they have to help them? What gauges can they use, other than their feelings and student feedback, that they are on the right track? In centrally driven programs, how do learning designers and their managers know what to refine and when to refine it?
We know that implementing new teaching strategies often takes multiple attempts for practitioners to refine and we also know that students’ perception of learning and their actual measured learning can differ dramatically. Our existing tools—the LMS, the digital textbook, and the webconference—often don’t provide practitioners with clear measurements and increased confidence that what they are trying to improve is, in fact, improving. It’s not magic to do so from a technical perspective. If you have good backward design, then you theoretically have good measures of whether, where, and how dramatically your teaching approach is succeeding or failing at getting students to succeed at the assessments that measure specific competencies. But our mainstream systems are not designed to make that easy. There is neither measurement nor reward for teaching in the way that CBE or even backward design requires to propagate. Without those functions, there is no positive evolutionary pressure on teaching. Regardless of how much educators may want to become better at their craft, their ecosystem is not conducive to that evolution.
And suppose an educator (or group of educators) does improve teaching of a particular topic. How does that spread? How does their Omicron become the dominant strain (to use a horrible analogy)? Right now, what happens in the classroom largely stays within the classroom.
The aforementioned evolutionary biologist Richard Dawkins wrote a seminal cognitive science article called “Selfish Genes and Selfish Memes.” While the viral pictures with captions on the internet owe their name to this article, he meant something broader. He proposed that an idea is like a gene. It wants to propagate. Because Dawkins’ “selfish gene” concept is a subtle one, let’s make our story a little simpler by thinking about an idea as functioning like a virus. (We do often talk about videos “going viral” on the internet.) Ideas “want” to spread to other hosts, where they can continue to establish themselves. Each host can be thought of as an ecosystem. For a literal virus, the relevant environmental factors include pre-existing immune responses, genetics, other health conditions that weaken the host, and so on. For a meme—an idea—the environmental factors that influence its success include the “infected” person’s strong conflicting (or supporting) ideas and values, prior interest in related ideas, and so on.
How do we enable better course designs, and better course design thinking, to propagate? How do we cultivate mindsets among educational professionals—the environmental conditions for the memes to take root—so that new designs are likely to spread? How do we cultivate continuous improvement and measurement as core values and part of educators’ sense of professional identity? How do we connect these ecosystems so that effective teaching approaches and course designs spread more easily?
We are not yet thinking clearly about how our ecosystems work. We are not yet approaching education as an organic, complex adaptive system. An ecosystem. And yet the biological principles apply whether we harness them or not because they are based on more fundamental rules about how feedback loops create stabiity and change. We ignore the basic mechanisms of evolution at our peril.
Reminder: This isn’t a book review or a position statement on CBE
My point isn’t to advocate specifically for this focus on CBE. I don’t believe that CBE is a panacea. Spoiler alert: Neither does Paul. On page 163 of Students First, at the end of several chapters of extraordinarily ambitious policy proposals, he writes,
Taken together, these recommendations for change are probably a bridge too far for the industry. Also, too much that is genuinely good about higher education might be placed at risk if we forced a sudden and dramatic shift to a completely competency-based education system.
So, the ask here is for something more modest. It is to create a safe space for innovation and learning to take place, to create a Plan B for those institutions willing to raise their hands and design new outcomes-based programs for students who want and need them.
Students First
(The book would have benefitted from Paul making this point clearly and loudly right at the beginning. That framing would have invited a more open-minded reading from skeptics.)
I have my own set of questions about CBE which I won’t recite here. I do agree wholeheartedly with Paul that we need a safe space for experiments, specifically for CBE and more generally for educational innovation. More than that, I think we should be creating conditions that reward bold experimentation, at scale, across a variety of learning contexts. And I think the only way we’re going to understand and mitigate the challenges of a change as dramatic as CBE while realizing its promise is to create an environment that encourages experimentation is by creating positive evolutionary pressure toward student success.
The real point of this post is that we should take the time and care needed to think deeply about how change happens in complex adaptive systems like education. We may very well be at a moment of dramatic change as external forces drive higher education from its balance point of punctuated equilibrium. Much is up for grabs now. As I wrote at the top of this post, the forces that determine whether a system has reached a point of stability don’t entail a notion of whether that point of stability achieves outcomes that are humane for individuals and productive for society. We need to develop and operationalize more robust theories of change. For the times, they are a changin’.
Saj says
When I need to understand something, educationally or otherwise, go to the experts eg Greta Thunberg has notoriety for her position on climate change but she is only the messenger, not the expert. When was the last time Paul taught a college course? I think we ought to listen to experts not messengers.
Michael Feldstein says
Paul runs a university of 180,000 students, many of whom take CBE programs and which gets solid results, particularly given the student population they serve. I wouldn’t call him a “messenger.” He could credibly be called one of the world’s foremost experts on running effective CBE programs at scale.
Daniel T Hickey says
Michael–
As always, for that deep objective dive into an important topic. I have followed LeBlanc’s efforts and SNHU over the years. I am quite fascinated by CBE and have a good friend who is finishing an IT degree at Western Governors University. As a scholar of educational assessment, I am going to take issue with your statement that once you define an outcome and assessment. “you can be pretty wildly experimental with how you teach. You are always measuring against results using the same yardstick.” That is simply not true because instruction can more or less aligned with the targeted assessment.
Take, for example, the field of cybersecurity education, where I have worked for a while. Consider, for example, a student who completes a high-quality cybersecurity degree, takes the Comp TIA Security+ exam because it is required for a job, and attains a passing score. Another individual lacking a degree might sign up for one of the numerous boot camps that advertise “free retake guarantees” (e.g., TrainingCamp.com), uses flashcards and other drill and practice methods to cram for the exam, and take the exam repeatedly until earning a passing score. Whereas the Security+ score for the degreed student represents just a sample of what the student learned across four years of coursework, the score of the crammer likely represents most of what the individual knows about cybersecurity. Furthermore, the degree program likely left behind knowledge that was more organized and structured and likely provided real practices with skills such as penetration testing and adversarial thinking. In contrast, the test prep likely left behind more isolated fragments of knowledge, much like the way that knowledge is represented on the exam. While the test prep likely taught ABOUT pen testing and adversarial thinking, it probably included no actual practice with these skills.
While some might attribute the preceding to the traditional test, that is a mistake. My careeer was shaped by my experience as a postdoc at the ETS Center for Performance Assessment. Between the time I accepted my postdoc in 94 and I got there in 1995, statewide K-12 “alternative” assessment reforms with performance and portfolio assessment disappeared seemingly overnight. Schools and states realized that reliability and validity plummeted, while costs skyrocketed and none of the promised instructional improvements were realized. Sam Messick (the late “grandfather of validity theory” at ETS) convinced me that construct irrelevant easiness (i.e., “teaching to the test”) was likely an even bigger problem with performance and portfolio assessment. This is because they are task-oriented (assessing what what someone did), rather than construct-oriented (measuring what someone knows about a domain). Sam warned of this problem in an essay that was published in Educational Researcher in 1984 entitled The Interplay of Evidence and Consequences in the Validation of Performance Assessments.
In response to concerns about workforce readiness, the cyber ed commuinity is investing massively in hyper-realistic “cyber range” simulations for both education and certification. Take three different undergraduates who perform equally well on a summative certification on the cyber range where they successfully identify and neutralize a threat in a complex real-world network simulation.
1. The first student was taught by an adjunct whose job depended on students passing that summative performance assessment and getting strong students course evaluations at the end of the course, after completing the assessment. That adjunct learns as much as possible about the summative scenario and spends the entire semester preparing students for it.
2. The second student was taught by a tenure track assistant professor who was also concerned with workforce readiness of the program’s graduates but still cared about passing rates and course evaluations. So that instructor prepared students more broadly but still made sure that the course prepared students for the certification task.
3. The third student was taught by a tenured professor who started by cybersecurity degree program and was only concerned with the program’s reputation for graduating students who are ready to secure nearly any network or platform. That instuctor did not even look into the cyber range certification scenario, had long since stopped reading course evaluations, and only cared about the occasional emails from employers expressing appreciation for another well-prepared graduate.
Despite the same scores, the third student clearly knows more than the second and a LOT more than the first. The same holds true if the instruction was carried out inside a range rather than a classroom. The situation is even more acute in portfolio assessments because the student in the first class would demand and presumably get extensive individualized feedback on interim artifacts.
I am NOT saying that this happens in all instances. I am impressed by the complex challenge projects and other curricula that my friend completed at WGU before passing Security+. But in my experience, many proponents of CBE are blind to the concern I raise here.
I look forward to reading Paul’s book. I am particularly interested in how he frames the equity response. CBE proponents (especially at the Aurora Institute) insist that “student agency” is the key to addressing system racism and historic inequalities, and Paul has placed equity prominently in his subtitle. But there are very few people within the culturally responsive pedagogy movement arguing for CBE. Quite to the contrary, it is difficult to see how CBE leads to anything but remedial “deficit-based” responses to educational inequities. I recently completed a pretty comprehensive review of the research and policy on CBE and equity. Let me see what Paul has to say and perhaps we can get a long-overdue conversation started here.
Michael Feldstein says
Thanks for the thoughtful and informative reply, Daniel. I take away a few points. First, if we’re trying to construct a useful independent variable in the form of a valid and reliable test of a real competency, we don’t yet have evidence that alternative assessments are better and, in fact, that they are often worse. Paul writes at length about the the challenge of valid and reliable assessments. I don’t think this is a contradiction but rather a challenge. Relatedly, if you’re going to keep traditional assessments, that tends to bring with it the problem of task-oriented teaching, or teaching to the tests, which lowers the value of the assessment. This is a hard problem.
With respect to the dependent variable—the teaching approach—as you know, the ability to “wildly experiment” is not the same thing as “anything you do is fine” in an experimental environment. It simply means that we can test different hypotheses about effective teaching practices. If it turns out that the traditional approach is the most effective one, that is what it is. It presents scaling problems for affordably reaching the students that Paul wants to reach but the point is to create an environment in which we can learn which teaching methods work how well under what sorts of conditions for which students.
I do think the conversation you call for at the end would be interesting and valuable, particularly if we can frame up the problem in a research-based context with a goal of arriving at new and better experiments for testing effective, scalable, and equitable approaches to education.