Why I host Learn Claude Code in NYC

I run a workshop series in New York called Learn Claude Code. People sometimes assume it is a marketing exercise. It is partly that, but the real reason is more selfish: teaching is the fastest way I know to actually understand a tool.

When you build alone, you route around your own confusion without noticing. Put twenty sharp people in a room with the same task and the friction becomes visible immediately. Where do they hesitate? Which mental model do they reach for that turns out to be wrong? What did I think was obvious that clearly is not?

The format

The sessions split into two tracks:

• Broad nights for newcomers, getting from zero to a working agent without the usual setup tax.
• Vertical labs for practitioners in a specific domain, finance, legal, private equity, education, who want to put agents to work on the problems they actually have.

The vertical labs are where it gets interesting. A lawyer and a quant ask completely different questions of the same tool, and both questions teach me something.

Why New York

New York has an unusual density of operators who are not engineers but are deeply technical about their own domains. They do not want a lecture about transformers. They want to know whether an agent can do the thing on their desk that eats three hours every week. That is exactly the right pressure to design against.

The compounding

Every session produces artifacts: skills, templates, patterns that worked and patterns that did not. Those flow back into the work I do the rest of the week. The community and the building are not separate activities. They are the same loop, run in public.

If you are in New York and want to come to one, reach out. The room is better with more perspectives in it.