Shawn
Wang

Shawn "swyx" Wang did not wait for permission to exist. He quit a $350,000-a-year quantitative finance job in his early thirties, taught himself to code on freeCodeCamp, and proceeded to become one of the most cited voices in AI engineering. That is not a rags-to-riches story. That is a thesis-to-proof story.

The thesis: that a new breed of engineer was emerging at the intersection of software development and large language models - not the PhD researcher training models from scratch, but the developer building products on top of them. The proof: a conference, a podcast, a newsletter, a phrase. "The Rise of the AI Engineer." When swyx published that essay, he was not describing a trend. He was naming it into existence.

Today, the AI Engineer Summit - which swyx co-founded with veteran conference organizer Ben Dunphy - is the largest technical AI conference for engineers in the world. From its inaugural 2023 event that sold out instantly with a 10-to-1 applicant ratio, it has grown to 6,000+ attendees, 29 tracks, 300 speakers, and four continents. The London leg alone drew over 1,000 engineers in 2026 and is planning for 2,000 in 2027.

But the conference is just one output. swyx is, above all, a prolific creator. He has published over 626 essays, tutorials, talks, podcasts, and notes. He co-hosts Latent Space, the top technical AI podcast and newsletter, with Alessio - a show that has reached 10 million readers and listeners through in-depth, in-person conversations with the people actually building AI at OpenAI, Anthropic, Meta, Databricks, and beyond. He founded Smol.ai, an LLM data pipeline company whose AI News newsletter is - fittingly - 99% generated by customizable research agents. He holds a GitHub repository of TypeScript cheatsheets with 47,000 stars. He wrote a 450-page handbook on developer careers. He helped define contributor guidelines for the React documentation.

All of this from a Singaporean who came to America for college and once spent his days trading currency derivatives in Excel and Haskell.

Before swyx was an AI engineer, he was a beginner. A finance guy who did not know React, teaching himself in public because he thought transparency might be useful to others in the same position. The essay "Learn in Public" that came out of that experience is now one of the most shared pieces in developer culture, read by millions and cited by people who have never met him as a turning point in how they thought about their careers.

The core idea is deceptively simple: create learning exhaust. Write the blog post you wish existed when you were stuck. Give the talk that would have helped you six months ago. Build the cheatsheet, start the newsletter, moderate the subreddit. Not for an audience - primarily for yourself, because that is what keeps you doing it. The audience is a side effect.

swyx practices this with an intensity that borders on systematic. He has a second brain in Obsidian. His ai-notes GitHub repository serves as both personal research notes and the data source powering Latent Space writing. He uses Twitter as a public note-taking tool. He practices what he calls "macro-tweeting" - resurfacing old tweets that aged well. His colleagues note that he is "right a lot." That is not luck. That is 626 published pieces compounding.

His personal operating system runs on three principles: Intentionality, Intensity, and Integrity. The Coding Career Handbook - his 450-page guide for junior-to-senior developer progression - is essentially those three principles applied to every stage of a software career, written for the person he was when he first opened a code editor and had no idea what he was doing.

Every industry has a moment where demand and supply discover each other and create an explosion. For AI engineering, swyx saw it coming before the explosion. Developers wanted to build with LLMs and did not know how. Companies needed engineers who could - not PhD researchers who spent years on model training, but practitioners who could ship products. There was no conference for these people. There was no clear name for them.

swyx fixed both. "The Rise of the AI Engineer" gave the category a name. The AI Engineer Summit gave it a gathering place. The 2023 inaugural event in San Francisco filled up within hours. The 2024 AI Engineer World's Fair drew 3,000 people to San Francisco - described as the largest technical AI conference for engineers in the world. By 2025, it had grown to 6,000+ attendees, 100 expo partners, and events on four continents including San Francisco, London, New York, Paris, Miami, Singapore, and Melbourne.

The 2026 London event (AIE Europe) exceeded 1,000 attendees. Planning for 2027 targets 2,000. The San Francisco event is eyeing Moscone Center. This is not a side project. This is a foundation.

The thesis that drove all of it - that there is a distinct and growing class of engineer building on top of AI rather than building the AI itself - has proven out. swyx also collaborates with Cognition on building rigorous evaluation standards for coding agents, pushing the field toward the kind of professional craft that sustainable engineering requires.

Latent Space - which swyx co-hosts with Alessio - is not another AI hype podcast. It is the one AI researchers and engineers listen to when they want to understand what is actually happening at the frontier. The show records mostly in-person, which is a deliberate choice. The conversations are technical, long, and specific. The guests are the people who built the thing being discussed - not commentators on the thing.

The show has interviewed engineers and founders from OpenAI, Anthropic, Google Gemini, Meta (including Soumith Chintala, the creator of PyTorch), Sierra (Bret Taylor), tiny (George Hotz), Databricks/MosaicML (Jon Frankle), Modular (Chris Lattner), and Answer.ai (Jeremy Howard), among many others. These are not surface-level conversations. They are the kind of technical deep-dives that turn a listener into someone who actually understands how Agents, Models, Infra, and AI for Science are being built right now.

Reaching 10 million readers and listeners is the kind of number that looks like it happened by accident. It did not. It happened because swyx has been creating learning exhaust at scale for years, and because the audience for serious AI engineering content is larger than most people thought.