What If the Three R's Aren't the Fundamentals Anymore?

Photo by Josh Applegate on Unsplash

Educators tell me repeatedly that the fundamentals don’t change. It is the most common response I get to any mention of sizable curriculum change. I’m told that reading, writing, and arithmetic—the “three R’s”—are eternal foundations that every generation needs.

I like to poke that bear. I ask them to consider how those fundamentals have historically changed, or to examine the massive shifts happening right now in how people actually consume and create information. I get pushback. “What the world values isn’t what we should value. Our standards are higher.” The fundamentals don’t change because we’ve decided they don’t change.

I’m not celebrating what’s coming. I worry deeply about AI’s power and who controls it. I worry about its use. But education systems can’t control that. They can’t drive the world in the direction they want through curriculum choices. That was true when schools formed during the Industrial Revolution because what schools taught aligned with what the world needed. When that alignment breaks, schools don’t reshape society through force of will. They’re along for the ride whether they admit it or not.

This article poses uncomfortable questions because pretending the world isn’t changing doesn’t protect students. It ensures they’re unprepared. There are educators who openly declare that their course content is eternally useful, that it’s good medicine, and then wonder why students go through the motions or fail to engage. Nothing is forever; this article asks whether “someday” might be an in-horizon timeline, plausibly visible rather than conjecture. Every schooling assumption needs to be on the table, quickly, because the world isn’t waiting for educators to feel comfortable.

If you’re already offended by my directness, then know I am not asking schools to abandon literacy and math. I’m trying to ask difficult questions, without really knowing what the answers should be. I just think the rocks are being pushed up increasingly steep hills.

The Three R’s Are Losing Utility Because They’re Becoming Optional

People have already been moving from text input and output to other modalities. Voice interfaces, video communication, audio content—these aren’t future technologies. They’re how people consume and share information now.

AI could replace much of the written layer and the mechanical or even conceptual math layers entirely. Want information from a document? AI reads and synthesizes it, presents findings verbally or visually. Need calculations? State your intent in natural language, AI determines what math applies and executes it.

If you don’t use a skill, you will lose it. This isn’t controversial; it’s biological reality. How good are most people going to get at any skill if they don’t have to use it much? The use-it-or-lose-it principle is already in effect.

Increasingly, schools aren’t going to make adults proficient at reading, writing, and arithmetic any more successfully than they’re going to ensure each adult knows how to grow their own food, repair their own car, or sew their own clothes. These were once considered essential life skills. Then systems changed. The skills didn’t disappear from humanity; they became specialized rather than universal.

I hate the calculator analogy for AI, but there are a couple of aspects worth highlighting. Even before calculators emerged, adults largely forgot the arithmetic they’d been taught in school. Many didn’t use the skill enough in daily life to be retained. We were filling a very leaky bucket. Plus, calculators didn’t actually worsen arithmetic skill. A 2003 meta-analysis of calculator use found no negative effects on students’ mathematical achievement, and research has found that children using calculators systematically throughout primary school actually had greater understanding and fluency with arithmetic. Mathematical reasoning survived, and perhaps was improved because more learning time could be devoted to such concepts. It’s probably true that people can’t do arithmetic in their heads as well anymore, but then again they don’t need to.

Clearly language will survive, including mathematical language. Yet people can look at the path toward AI doing all of the software coding and realize that the need to teach coding languages is also much reduced, but not see the same progression with other texts. I certainly don’t think the conceptual principles or thinking approaches behind coding are going anywhere, nor should they. Ditto for abstract thinking and conceptual understanding regarding language and math. And human languages are of course spoken too; they’re not going anywhere. The question is whether textual encoding of language and symbolic representation of mathematical concepts need to be universally mastered skills, or whether they become specialized capabilities.

Literacy and numeracy serve multiple purposes—they’re ways of representing information, methods of archival, and when learned deeply, they become thinking tools themselves. But perhaps those same goals can be met through other information forms now. If that’s true, maybe reading and writing can be learned to a substantial degree at a less intensive pace, thereby allowing room for topics that require more emphasis in the changed world, as discussed in the next section.

Some have called this a return to an oral culture, but I don’t think that’s correct. Historical oral cultures were limited by sparse archival systems that constrained building systematically on accumulated knowledge across generations. Written language created civilizational acceleration precisely because it preserved and transmitted insights reliably. The future, however, is something entirely new. AI provides all the benefits of written archival and knowledge transmission without requiring every human to encode and decode text manually. Humans can create and consume through image, video, voice, and interactive media while AI handles the textual layer, the mathematical layer, the archival function.

Many brains aren’t set up for literacy and numeracy anyway. These aren’t evolutionary features—human brains didn’t develop to decode arbitrary symbols or manipulate abstract numerical representations. Dyscalculia affects 3-7% of the population, while dyslexia affects 20%. Those whose brains struggle with these artificial skills get labeled “special needs.” Some of these struggling students might excel at pattern recognition, systems thinking, spatial reasoning—capabilities the current paradigm undervalues because it’s fixated on textual and numerical proficiency.

If people aren’t reading in a decade, should schools still teach reading as a fundamental? Maybe reading becomes specialized—taught later, taught alongside other modes, taught to those who need it. Maybe literacy being the gateway toward higher-order thinking only was true when information was scarce and delivered textually.

Human brains weren’t designed for literacy and numeracy. Making brains literate requires years of intensive, often painful training. That training was justified when text was the only gateway to information and higher-order thinking. It’s not anymore.

Durable Skills Are More Important and They Don’t Require Literacy First

For decades schools have touted the importance of durable skills—critical thinking, collaboration, creativity, problem-solving, etc.. But the reality is they’re not well set up to teach them. Most schools operate through bottom-up, behaviorist approaches: drill procedures, assess mastery, move to the next level. As explained in my first book Wisdom Factories, durable skills aren’t learned that way. They’re learned intuitively through experience and reflection. You don’t master judgment by memorizing decision frameworks; you develop it by making decisions, seeing consequences, and refining your approach. If durable skills matter—and everyone claims they do—why not actually put them front and center instead of treating them as supplements to mechanical mastery?

Children can demonstrate durable skills at very young ages, and these capabilities can be developed in age-appropriate ways. Preschoolers demonstrate reasoning and problem-solving. Four-year-olds evaluate information sources and distinguish accurate from inaccurate ones. Young children understand deception, make fairness judgments, and solve problems creatively. They’re thinking critically about their world. It’s not necessary to wait until some defined cognitive developmental threshold. Durable skills are relevant at every age.

The critical learning period argument cuts both ways. Yes, certain capabilities become harder to acquire with age. Language learning gets more difficult after early childhood. Musical abilities develop most naturally when started young. If durable skills also have sensitive developmental periods—and there’s every reason to think they do—then spending those critical early years drilling mechanical skills means missing the windows when sophisticated thinking capabilities develop most naturally. Students who struggle with literacy might excel at pattern recognition, systems thinking, interpersonal judgment. The current paradigm doesn’t emphasize those capabilities because it’s fixated on textual and numerical proficiency as prerequisites.

There are particular durable skills related to understanding and using AI that were laid out in AI Wisdom Volume 1 and (soon to come) Volume 2. Like most durable skills, the foundations are meta-skills, in this case principles about how intelligence works, how systems behave, how to orchestrate capabilities rather than execute tasks. Do you need to learn the nuts and bolts of something to develop meta-skills? Sometimes, but often they’re completely disconnected areas. Knowing how to fix a car engine tells me very little about how to design a car or drive one safely. Learning to orchestrate AI capabilities doesn’t require mastering the mechanical literacy AI will handle. The increases in the abstraction level of human contributions that AI accelerates also in many cases means learning entirely different knowledge not super dependent on the lower abstraction levels.

---

Does AI remove the last reasons to maintain universal literacy and numeracy? Information doesn’t need textual transmission. Math can be done from intent alone. AI can interpret, synthesize, and present findings without humans reading source material.

The “you need literacy to check AI’s work” argument assumes humans actually use deliberative processes to validate information. They don’t. People rely on intuitive pattern recognition and quick heuristics because information volume makes deliberation impossible. AI can validate other AI’s work with confidence scores and multiple independent checks. That’s likely more reliable than humans trying to deliberate through everything they can’t process anyway.

This isn’t 20 years out. Students entering elementary school now, or maybe even middle school, will graduate into a world where the norm might be that AI does the reading, writing, and math. Voice interfaces, multimodal AI, and visual information processing will all dominate information access and creation. Economic bubbles don’t erase technological capability. Even if AI development slows economically, the core capability isn’t disappearing, it’s accelerating.

In the end, does AI do ALL the writing and ALL the math, while humans consume and create through whatever modes they prefer—voice, video, visual, in-person? Text and numerical symbology become internal languages for machines, not skills every human needs?

The capability exists now. Text becomes optional, used when convenient but not required. I can think of a ton of reasons that would be bad, and sad, but I can’t think of many reasons it isn’t plausible. If students also see that future, and I think some see it better than many educators, then expect learning disengagement to amplify.

The educational system shaped society when curriculum aligned with what the world needed. When alignment breaks—as it’s breaking now—and when AI is an alternative source of knowledge and training, then schools and colleges lose relevance. This is no different conceptually than when an AI conversation goes off the rails from your intent. You can try to bring it back into alignment, but if it has diverged too far you’re better off starting a new conversation. For a student, whether a child or adult, that “new conversation” might be learning from AI because they know they won’t get it from formal education.

I’m not trying to dissuade the teaching of reading, writing and math. That’d be tilting at windmills anyway. I am trying to get you to frame it differently. The three R’s have always been a vehicle to deeper thinking, thinking that could only be unlocked via text modes because other forms weren’t available or accessible, such as talking to actual human experts. That’s not true anymore. We’d better start working through the implications.

Schools either adapt or become gatekeepers to skills that matter less each year. Students will inherit a world where AI handles mechanics and human wisdom guides outcomes. I’d say focus on the wisdom part. Otherwise, it’s like promoting the best worker to be the boss. Don’t expect the skills to transfer to an entirely different role.

Increasingly, teaching that wisdom doesn’t require the three R’s.

©2025 Dasey Consulting LLC