I’m going to say something that might get me in trouble with my friends in Silicon Valley. Most of what you’ve been told about teaching AI is wrong. Dead wrong.
I’ve seen it from every angle. As a founder who built and sold two tech companies, one of which was acquired by Gusto. As an angel investor who has written checks to over 200 startups, including some of the biggest names in AI like Anthropic, OpenAI, and Scale AI. And, believe it or not, as someone who has a deep passion for education.
What I see in most AI curriculums makes me cringe. They’re usually just a laundry list of the hottest tools and frameworks. “Learn Python, then TensorFlow, then PyTorch, then this new shiny thing that just came out last week.” It’s a recipe for creating technicians, not thinkers. It’s like teaching someone to be a chef by only giving them a list of kitchen appliances.
We’re preparing students for a world that will look nothing like the one we live in today. The tools will change. The platforms will evolve. The only thing that will remain constant is the need for critical, independent thought. And that’s what we should be teaching.
The Fallacy of the “Tool-Based” Curriculum
I once sat in on a curriculum planning meeting for a university’s new AI program. The excitement in the room was palpable. They had a whiteboard filled with every buzzword you could imagine: “Deep Learning,” “Natural Language Processing,” “Reinforcement Learning,” “GANs.”
I listened for an hour as they debated whether to teach TensorFlow or PyTorch first. I finally had to raise my hand. “What problems are the students going to solve with these tools?” I asked.
Silence.
They were so focused on the “what” that they had completely forgotten the “why.” This is the fundamental flaw in how most people approach AI education. They think that by teaching the tools, they are teaching AI. They’re not. They’re teaching a set of instructions that will be obsolete in a few years, maybe even a few months.
Think about it. When I was starting my first company, MovieLaLa, the AI landscape was completely different. We were using collaborative filtering algorithms to build a movie recommendation engine. Today, you can do that with a few lines of code from an open-source library. If I had only learned the specific tools of that era, I would be a dinosaur today.
My Counterintuitive Approach: The Three Pillars of an Effective AI Curriculum
So, how do we do it right? How do we create a curriculum that prepares students for the future, not just for the next job interview? I believe it comes down to three core pillars.
Pillar 1: Start with Problems, Not Tools
Instead of starting with a list of technologies, start with a list of problems. Real, messy, interesting problems. For example:
- How can we use AI to detect wildfires earlier?
- How can we build a tool to help people with dyslexia read faster?
- How can we create a more personalized learning experience for students?
When you start with a problem, the tools become a means to an end, not the end itself. Students are forced to think critically about which tools are right for the job. They learn to be resourceful, to experiment, and to fail. That’s a much more valuable skill than knowing the syntax of a particular library.
I remember when the team at Scale AI, one of my early investments, was just starting out. They didn’t have a perfect, pre-defined tech stack. They had a massive problem: how to get high-quality training data for AI models. They experimented with different approaches, built their own internal tools, and iterated relentlessly. They were problem-solvers first, and technologists second. That’s why they’re a decacorn today.
Pillar 2: Build a Foundation of Core Concepts
While I’m not a fan of tool-based curriculums, I do believe in building a strong foundation of core concepts. These are the timeless principles that will be relevant no matter how the technology changes. I’m talking about:
- Statistics and Probability: The bedrock of machine learning. You can’t understand how a model works if you don’t understand the math behind it.
- Linear Algebra: The language of data. Everything in AI, from a simple regression to a complex neural network, is built on linear algebra.
- Computer Science Fundamentals: Data structures, algorithms, and computational complexity. These are the building blocks of any software system, AI or otherwise.
- Ethics and Critical Thinking: This is the most important, and most often neglected, part of an AI education. We need to teach students to think critically about the societal impact of AI, to ask hard questions about bias, fairness, and transparency.
I’d rather hire an engineer who has a deep understanding of these core concepts and has to learn a new framework on the job, than someone who knows all the latest tools but can’t tell me the difference between a p-value and a confidence interval.
Pillar 3: Foster a Culture of Continuous Learning
The only certainty in the world of AI is that things will change. The most important skill we can teach our students is how to learn. How to stay curious, how to adapt, and how to embrace the unknown.
How do we do that? By creating a learning environment that mirrors the real world. That means:
- Project-based learning: Give students long-term, open-ended projects where they have to define the problem, choose the tools, and manage the entire process from start to finish.
- Collaboration: No one builds anything meaningful in a vacuum. Students need to learn how to work in teams, how to communicate effectively, and how to give and receive constructive feedback.
- Connecting with the industry: Bring in guest speakers from startups and established tech companies. Have students work on real-world projects with industry partners. The more exposure they have to the real world, the better prepared they will be for it.
A New Manifesto for AI Education
So, here’s my challenge to every educator out there. Throw away your old curriculum. Stop chasing the latest fads. Start with the problems. Build a strong foundation of core concepts. And create a culture of continuous learning.
Let’s stop training technicians and start cultivating the next generation of AI leaders. The ones who will not just use the tools we give them, but will build the tools of the future. The ones who will not just solve the problems we assign them, but will identify the problems that we don’t even know exist yet.
It’s a taller order, I know. It’s much easier to create a checklist of tools and call it a curriculum. But our students deserve better. The future demands it.
I’ve made 200 angel investments. I’ve seen what it takes to build a successful company from the ground up. And I can tell you this with absolute certainty: the most valuable people in any organization are not the ones who know the most about a particular tool. They are the ones who can think critically, solve complex problems, and learn new things quickly.
Let’s create an AI education system that produces more of those people. The world will be a better place for it.
Frequently Asked Questions
How do I measure success with this approach?
Pick one or two metrics that directly tie to your goal and track them weekly. Vanity metrics like page views or follower counts rarely matter. Focus on metrics that reflect real engagement or revenue impact.
What tools do I need to get started?
Start with the basics. You don't need expensive software or fancy tools. A spreadsheet, a note-taking app, and direct access to your customers will get you further than any enterprise platform. Add tools only when you hit a specific bottleneck.
What are the most common mistakes when designing an ai curriculum (the counterintuitive guide for educators)?
The biggest mistake I see is overcomplicating things early on. Start with the simplest version that works, get real feedback, and iterate from there. Another common trap is copying what worked for someone else without understanding the context behind their decisions.