A curriculum for the AI age

Designing an Education System for the Last Human Century

I’m Lawrence. a pleasure. I invest in entrepreneurs making the world better for my children lawrence@lunar.vc.

---

I've explored how AI is already automating knowledge work—including literally all my research tasks. This raises a question: Why are we still educating children for an economy that won't exist when they graduate?

I have done more than thought about this. This isn’t another VC opining on another thing HE knows nothing about. Welcome to the Internet. Or the US Government now I guess.

No, my wife and I moved across the country away from friends and family because we truely believe the education system is broken. I am in-all. We pulled the lever and diverted the trolley. Why?

For Whom the Bell Tolls

AI is our last invention. For the impact on education I don’t have to be specific on timlines. Do you think AI will automate most of the cognitive tasks by 2028? 2030? 2035? Fine, for our purposes, education is a 15-year haul. So we have until 2040. Do you think AI progress is still gonna stop? I suggest you get off at the next stop. Copium is a hell of a drug.

Our education system was never really designed for human flourishing. Soz to disappoint. There was always a tension between industrial efficiency and human development. It’s not entirely true that “education is to create workers for the factory”. I mean the entire American education system was based on John Dewey’s ideas of a child-centred approach and democracy in education. But also, the truth is, the Prussian model that forms the backbone of modern schooling had a specific purpose: to produce standardized workers for an industrial economy. From the playground to the factory floor.

Look at a typical school structure: Age-based cohorts? Batch processing. Bell schedules? Factory shifts. Standardized testing? Quality control. Emphasis on compliance? Workforce preparation. Credentials? Employer sorting mechanisms. We were institutionalised. Take a step back and think how mental it is.

Shadow a typical kid in a secondary school: Seven disconnected subjects. Bells dictating when thinking starts and stops. Oh you were busy learning about algebra were you? Too bad, pack up and off you go to Religious Education now. Quick snap and straighten up your tie. Fragment your thinking, respond to authority, prepare to be measured. No wonder school success predicts conscientiousness—following rules without questioning their fundamental purpose.

Despite decades of supposed reforms, this fundamental architecture remains stubbornly intact. We've added computers and buzzwords like "critical thinking" and "21st-century skills," but the baseline assumptions persist. We're systematically preparing students for obsolescence—a steepening ladder of achievement while the economic ground beneath it rapidly becomes peformatuive and devoid of meaning.

What Remains Human?

I explored this a little in the end of data driven VC essay, what will all the VCs do? Why does noone ever think of the VCs! Less importantly, however is what will the other less imortant humans do? Not in information processing or pattern recognition. Not strategic planning. AI is coming for that. What will be left are things that humans want other humans to do.

1. Authentic Presence - We desire genuine emotional connection, especially in contexts of trust, vulnerability, or grief. A patient wants their doctor to truly understand their fear; a grieving person needs authentic compassion, not simulated empathy. No matter how sophisticated AI becomes at mimicking emotional intelligence, we can sense the difference between algorithmic approximation and genuine human care.

2. Accountability - We expect moral responsibility—someone to take ownership of decisions, not just simulate ethical reasoning. When difficult trade-offs must be made, we want decision-makers who can be held accountable, who feel the weight of their choices, and who understand the lived experience of those affected. AI can weigh variables but cannot truly bear responsibility.

3. Experiential Intuition - We trust judgment grounded in lived cross-domain experience when navigating ambiguity, especially in leadership, caregiving, or mentorship. This isn't mere pattern recognition but a form of wisdom that integrates cognitive, emotional, and ethical dimensions through the substrate of personal experience. Leaders who have weathered failure possess something algorithms cannot: embodied knowledge of what it means to face human challenges.

4. Cultural Creation - This is a good one, and one we should think about more as it relates to creativity. We value human participation in generating meaning, art, ritual, and community—expressions that reflect our shared humanity. While AI can generate content that mimics human creativity, it cannot participate in culture as a genuine stakeholder with its own existential concerns. Human art matters not just for its output but because it represents authentic expression of human experience.

These aren't tasks humans perform better—they are roles humans inhabit more meaningfully. The distinction isn't about efficiency but about authenticity.

Human-Centered Alternatives

The answer is not more screens, YouTube and AI tutors. Yes personalised learning is exciting. I get it. I’m using Claude for flashcards too. But sadly for me, as an investor, the answer is not EdTech. Adding more technology to an outdated educational model only digitizes its flaws. Instead, we should look to educational approaches that already cultivate the distinctly human capacities that will remain valuable in an AI-dominated future:

1. Montessori education develops internal authority rather than dependence on external validation. Students learn to identify meaningful questions, design their own learning paths, and cultivate intrinsic motivation—the drive to learn for its own sake rather than for grades or approval. In Montessori classrooms, you'll find children of different ages working at their own pace, deeply engaged in self-chosen activities, with teachers serving as guides rather than instructors. This approach directly nurtures authentic presence by teaching children to recognize and follow their genuine interests rather than performing for external rewards. It cultivates ethical accountability by giving children real responsibility for their learning environment.

2. Waldorf education develops synthetic rather than merely analytical thinking through educating "the head, heart, and hands." Students engage with material through multiple modalities—intellectual understanding, emotional connection, and physical creation. A Waldorf classroom might explore geometry through drawing precise forms, discussing their cultural significance, and building three-dimensional models. This integrated approach develops experiential intuition by connecting abstract concepts to emotional and physical experiences. It fosters cultural creation through its emphasis on arts, storytelling, and celebration of cultural traditions.

3. Forest schools develop embodied intelligence through extended time in natural settings. Children spend most of their day outdoors, learning through direct interaction with the natural world. A forest school session might involve building shelters, tracking animals, or navigating terrain—activities that develop risk assessment, problem-solving, and adaptability. This approach directly cultivates physicality by engaging the whole body in learning. It builds comfort with uncertainty by immersing children in complex natural systems that cannot be fully controlled or predicted. Forest school graduates develop the kind of embodied wisdom and resilience that will remain distinctly human even as AI masters abstract problem-solving.

These approaches aren't just theoretical alternatives—they're proven educational models that already prioritise the human capacities AI cannot replicate. But how many graduates went to Oxbridge you ask? Get out…

A New Curriculum

Drawing from these alternatives, education optimized for 2040 would be centred around:

1. Purpose - Purpose-finding must become central to education because AI is rapidly eliminating the traditional sources of human identity found in work. Unlike knowledge acquisition (which AI excels at), purpose is fundamentally human. As automation renders many careers obsolete, the ability to derive identity from internal values rather than external roles becomes not just fulfilling but necessary for psychological survival.

2. Wisdom - While AI can process vast information, wisdom—the contextual application of knowledge—remains uniquely human because it requires integrating cognitive, emotional, and ethical dimensions simultaneously. AI can simulate ethical reasoning but cannot truly weigh values across domains without human guidance. This distinction matters because algorithmic recommendations without wisdom lead to technically correct but humanly disastrous outcomes.

3. Empathy - Human social intelligence represents not just a competitive advantage but an entirely different category of cognition from what AI can achieve. While AI can recognize emotions, it cannot genuinely feel them or understand their experiential quality. This distinction creates an unbridgeable gap in social contexts where authentic emotional resonance matters. When researchers attempt to simulate empathy, they reach a computational barrier—the "empathy gap"—where models can mimic but not truly experience the internal states of others.

4. Physicality - Our embodied nature is fundamental to human intelligence in ways AI fundamentally cannot replicate. One day, robots etc. Embodied cognition research demonstrates that physical experience shapes conceptual understanding through motor resonance and experiential metaphors. When we navigate physical space or manipulate objects, we develop intuitive physics and spatial reasoning that ground abstract thinking. This physical dimension of intelligence creates a competitive advantage in domains requiring adaptive interaction with unpredictable environments. (for now)

Can a new educational setting, microschools, homeschool hubs, or even traditiinal schools design a curriculum around these 4 skills?

The Key: Breaking the Link Between Identity and Productivity

Much of this isn’t new. Dewey, Montessori and Steiner were knocking about in the 1910s. But there is an urgency now that hasn’t yet been grasped.

The future of education isn't about preparing for jobs AI will eliminate or hand-waving and saying new jobs emerge like "VR designer" or "drone delivery driver." It's about preparing humans for what remains fundamentally human: finding purpose, building relationships, exercising wisdom, and creating meaning in a world where economic productivity no longer requires most of our cognitive capacity. Yes, it's "woo-woo"—because the most human elements of existence always seemed that way to industrial thinking.

The only way we can structurally address AI automation is to prepare our children for a world where they don't tie their identities and self-worth to productivity and a job. The automation of knowledge work doesn't make education obsolete—it makes human-centered education more essential than ever. Our challenge isn't technological but philosophical: Can we let go of industrial values deeply enough to prepare young people for a world where being distinctly human matters more than being economically productive?

---

Thanks to Jonatan for telling me my first draft was bad. I didn’t send him this latest draft, in case he said this was bad too. Oh yeah, there are many drafts of these essays. I don’t just bang them out in 10 minutes, even if it reads like I do. It’s a bit.