The Gasp: NYU, March 2025
Fifth in a series about my AI lecture tour through Ivy League and international universities
Nerd to Nerd
NYU was almost all engineers.
I stayed at the Moxy on the Lower East Side - modern, minimal, the kind of hotel that doesn’t try too hard. Perfect for the vibe of this trip.
After the mixed audiences at Harvard, the business students at MIT Sloan, the international blockchain crowd in Amsterdam - NYU was pure. Computer science students. Future developers. My people.
A nerd talking to nerds.
The Down and Dirty
I could go deep here. Really deep.
Cursor deep dive - Not just “it’s an AI code editor.” I showed them the actual workflow. How to use it. What it’s good at. Where it falls down. The prompting techniques that make it actually useful.
Backend breakdown - ChatGPT vs Claude vs Gemini. Not the marketing. The actual behavior differences. When to use which. The personality quirks. The failure modes.
Open source models - What you can run locally. What it means for privacy, for experimentation, for building things you actually own. LM Studio, Ollama, the ecosystem.
The DeepSeek moment - This was happening right around then. DeepSeek R1 dropped and everyone was trying to figure out what it meant. Chinese model. Open weights. Reasoning capabilities. A fraction of the cost.
I could talk about all of this because the audience could follow it. They had the context. They had the vocabulary. They wanted the technical details.
The Gasp
Partway through the agentic workflows section, I stopped the prepared material.
“I want to be direct with you about something,” I said. “Pretty soon, you’re not going to have to write a single line of code.”
The room gasped.
An actual, audible gasp. From a packed room of CS students.
Then everyone started talking to each other. Murmuring. Processing. They couldn’t believe what they’d just heard.
I rattled them.
The Point
I wasn’t trying to scare them. I was trying to prepare them.
The direction this is heading - reasoning models, agentic frameworks, natural language as the interface - means the role of the developer fundamentally changes.
Code monkeying goes away. You don’t need humans to type syntax anymore. The models do that.
What you need is people who understand systems. Who can architect solutions. Who can think about what should be built, not just how to type it.
The engineers who understand the larger concepts - they win. The ones who only know how to bang out code - they’re in trouble.
That’s the message. Not “your skills don’t matter.” But “your skills need to evolve.”
The Hunger
These kids were hungry.
Not scared. Hungry.
After the gasp settled, the questions came fast. Not “are you sure?” defensive questions. But “how do I get ahead?” questions. “What should I be learning?” “Where’s the opportunity?”
They got it. The world is changing. They wanted to be ready.
That energy - young engineers who see transformation coming and want to ride it rather than fight it - that’s what keeps me optimistic about the future.
The Tools in March 2025
For the record, here’s what the landscape looked like:
Reasoning models: o1/o3 from OpenAI, Claude’s extended thinking, DeepSeek R1 just dropped Code editors: Cursor leading, Copilot improving, new players entering Agentic frameworks: LangChain, CrewAI, Swarm, AutoGen - everyone building No-code platforms: Zapier, Make, n8n - connecting everything Local models: Ollama, LM Studio - run your own
The pace of change was ridiculous. Something new every week. The students were trying to keep up. I was trying to help them see the pattern behind the chaos.
The Feeling
NYU was glorious.
That’s the honest assessment. A room full of smart, hungry engineers, ready to go deep, ready to be challenged, ready to hear the truth about what’s coming.
The gasp. The conversations. The energy.
That’s what teaching should feel like.
Next: Amsterdam - an international crowd at a blockchain conference where I talked about robots.
Date: March 6, 2025 Venue: NYU Audience: Almost all engineers Key moment: “You won’t have to write a single line of code” → audible gasp Vibe: Nerd to nerd, hungry crowd