Profile
The Man Who Couldn't Stop Reading About AI
Charlie Guo is now a Developer Experience Engineer at OpenAI, where he helps builders around the world figure out how to actually use the most powerful AI models ever made. His job is to turn complexity into clarity - through documentation, guides, tutorials, and worked examples. It sounds simple. It is not. And Charlie is unusually good at it.
The path from curious programmer to OpenAI employee reads less like a career plan and more like a series of bets on personal obsession. He read dozens of AI articles before going to sleep at night. He started a newsletter to process what he was learning. He tinkered. He published. He tinkered more. The newsletter grew to 23,000 subscribers. OpenAI noticed. None of this was inevitable. All of it was consistent.
Before landing at OpenAI in February 2026, Guo spent years at the intersection of software and startups - as a co-founder, as a CTO, and more recently as a Staff AI Engineer at Pulley, the cap table management company. Through all of it, one thing stayed constant: he was building in public, writing in public, and learning in public, usually before anyone thought there was a career to be made from it.
"Thinking about this technology has kept me up at night."
- Charlie Guo
🎓
Stanford, 2013
While still an undergrad studying CS, Charlie co-founded ClassOwl - an edtech startup partnered with Stanford to improve student-teacher communication. Architecture built in Python and Django. Later sold to Branch Metrics by his co-founders while he moved on to the next thing.
1
🚀
YC: FanHero
Post-Stanford, Guo launched FanHero - a startup that got accepted into Y Combinator, the Silicon Valley accelerator that has backed Airbnb, Dropbox, and Stripe. He's also been backed by 500 Startups and StartX. Three accelerators before most people have one.
2
✍️
The Newsletter Era
When ChatGPT launched, Guo had "minimal AI knowledge" by his own admission. He started building projects and writing about them. Artificial Ignorance began not as a media strategy but as a learning log. Three years later: 23,000 subscribers and an OpenAI offer.
3
The Work
Artificial Ignorance: What It Is, Why It Works
The newsletter is called Artificial Ignorance, and the name does most of the explaining. It lives at ignorance.ai - a domain that is either inspired or absurd depending on your relationship with irony. It's a Substack publication that covers AI at the intersection of engineering and business, with a specific audience in mind: software engineers, product builders, and startup founders who want to get their hands dirty with AI without drowning in hype or jargon.
Michael Spencer, writing for the AI Supremacy newsletter, describes it as "one of the clearest newsletters around with insights you cannot find anywhere else." That's the thing about Charlie's writing - it's not hot takes dressed as analysis. It's closer to what happens when a senior engineer sits down to explain something they've actually worked through themselves.
The newsletter went from around 5,400 free subscribers to over 23,000 in roughly two years. It reached Substack bestseller status. The growth strategy was not aggressive distribution tactics - it was consistency, collaboration with other newsletter writers, and the slow compounding effect of writing things that are genuinely useful. Charlie tried podcasting and live workshops. Neither stuck. He abandoned both rather than forcing them. A useful reminder that knowing what to stop is as important as knowing what to continue.
His self-described method: "Read, think, tinker, write, repeat." Five words. No MBA required.
The Tech Side
Self-Taught, High-Output, Deeply Opinionated
Here's the thing about Charlie Guo the engineer: he has fifteen years of programming experience, but he taught himself AI engineering from scratch after ChatGPT launched. He started from zero. He built a voice-based smart assistant. He built a "chat with your documents" app. He built a fake nature documentary generator - purely because it was interesting and because building something absurd is one of the fastest ways to learn something real.
His GitHub repository charlierguo/gmail - a Pythonic interface for Google Mail - dates back to 2013 and has accumulated 1,800+ stars. It's a decade-old library that still gets attention, which tells you something about the quality bar he sets even for side projects. He's an Arctic Code Vault Contributor, a distinction GitHub awards for contributions that are literally preserved for future civilizations. Not bad for someone who also writes a newsletter.
At OpenAI, Guo sits on the Developer Experience team - the group responsible for making sure that when a developer picks up OpenAI's technology, they don't immediately get lost. This is harder than it sounds. The audience ranges from beginners who've never called an API to experienced ML engineers who want to push the limits of what's possible. The job is to write for all of them, clearly, without talking down to anyone.
"To understand what AI is capable of, there is no substitute for working with it yourself."
- Charlie Guo
Backstory
From Crowdmade to OpenAI: The Long Game
The decade between Stanford and OpenAI involved more pivots than most resumes admit to. After FanHero and ClassOwl, Guo co-founded Crowdmade, took a stint as CTO of an eCommerce company, eventually landed at Pulley as a Staff AI Engineer, and all the while kept writing Artificial Ignorance. He also found time to write a book - Unscalable, a collection of startup interviews and lessons about "doing things that don't scale." The book is available through AbeBooks with an ISBN that suggests it actually exists in physical form, which in 2024 feels like a commitment.
His writing has appeared in Wired magazine, which is less surprising once you've read a few issues of Artificial Ignorance. He writes the way someone writes when they've internalized the material well enough to say what they actually think, rather than what sounds safe to say.
When he announced he was joining OpenAI in January 2026, he described the path as "fairly unexpected." That's probably true and also probably undersells how directly his three years of public writing about AI led to the offer. Thinking in public is how ideas become credentials. Charlie Guo is an unusually clear example of how that works in practice.
Three accelerators. Three startups. One newsletter. One OpenAI offer. The math checks out - eventually.
Character
Hype-Free Zone, Population: One
Read enough of Artificial Ignorance and a personality emerges. Guo is the kind of person who experiments before he opines, who abandons projects that aren't working (the podcast, the workshops), and who is genuinely more interested in what AI can actually do than in what it might theoretically do in some speculative future. He is, by his own description, deeply curious about the technology in a way that has kept him up at night. That's not a metaphor. He reads about AI before bed. He has for years.
His writing style reflects this: no superlatives, no predictions cloaked as facts, no AI hype translated into AI anti-hype. Just: here's what I tried, here's what happened, here's what it means for you. His endorsers - other newsletter writers in the AI space - consistently reach for the word "clear" when they describe his work. That's the highest compliment in a field that rewards the appearance of depth over actual depth.
He's also patient in ways that are genuinely rare. He published consistently for three years before the newsletter became a meaningful career asset. He learned AI engineering in public before there was a job title for it. He's not in a rush, except perhaps when it comes to understanding what new AI capabilities actually mean. On that, he moves fast.
In His Own Words
What Charlie Guo Says
To understand what AI is capable of, there is no substitute for working with it yourself.
There's no established curriculum or career track for building with foundation models.
Read, think, tinker, write, repeat.
Thinking about this technology has kept me up at night.
