I want to be precise about what this post is and what it is not. It is not "anyone can do this." It is not motivational content. It is an honest accounting of what the credential-free path actually looked like — what it cost, what it built, and what I would do differently.
I earned my GED in juvenile detention. That's the real sentence. Not "I dropped out" — I was locked up, and I sat in a facility and passed a test, and that became the highest formal credential I have ever held. I wrote a whole book about what that period was and what it made me. The short version is: it was the place that broke some things and hardened others, and somewhere in it I decided I wasn't going to let the ceiling other people could see actually function as a ceiling.
The GED from juvie is a specific kind of weight to carry into rooms. Not impossible — but not nothing either. Every pitch, every partnership conversation, every moment where someone asks about your background. The credential is a proxy for competence, and the proxy is broken, but it is still the currency a lot of gates require.
What I figured out is that you can route around the gates. Not by ignoring them — by building something that makes the question irrelevant. When you own the company, nobody asks for your transcript.
I want to be accurate here because I see a lot of content that either oversells or undersells this. AI tools didn't build things for me. They built things with me — and the ratio shifted over time as I understood the underlying systems better.
The mistake people make is treating AI assistance as a substitution for learning. It is a compression of the learning curve. You still have to understand what you built — because when it breaks at 2 a.m. and a lead pipeline is down, there is no AI tool that can care about it the way you do. You have to know the system. The difference is how fast you get there.
The lonely part. Nobody tells you about the lonely part.
Self-education is not a schedule. It is not a cohort. There is no professor to email, no study group, no graduation ceremony. There is just you, a problem you don't understand, and the decision to keep going or not. The decision is made in small increments, late at night, when the thing still doesn't work and you've been at it for four hours and every instinct says to quit and get a real job with a real salary and let someone else carry the weight.
The tools help. But the tools do not replace the decision.
Legitimacy, mostly. The credential. The right to be in certain rooms without having to explain yourself first. Access to networks that require a handshake I don't have. These are real costs of not having the degree. I'm not pretending they aren't.
What it would not have bought: the specific knowledge I have. The architecture patterns I discovered by breaking things. The business instincts that come from losing money and understanding why. The writing voice that came from publishing 18 books without anyone telling me I had to write them a certain way. Those things are the product of constraint, not curriculum.
You learn differently when failure is direct and immediate. A bad system prompt in production is down now — not a grade, not a note in the margin. You learn differently when the company you are building is the company that pays you. There is no gap between theory and application. They are the same thing.
Find peers earlier. The isolation of self-teaching is partly chosen and partly structural — the formal credential systems have built-in cohort networks and I built mine later than I should have. I spent too long solving problems alone that people around me had already solved.
Document the learning faster. The 18 books are partly a function of this realization: every time I figured something out, I wrote it down. Not because I thought people would read it, but because the act of writing it forces you to understand it. The books came from the documentation, not the other way around.
Trust the compounding earlier. Three months in, I could build a basic Cloudflare Worker. I couldn't see the 87-worker architecture from there. Six months in, I could see it. Eight months in, I'd built it. The compounding is not visible at the beginning. You have to trust the direction and keep moving.
The $0 education is a real number. But the real cost was time, and that cost was not zero — it was the highest-cost item on the list. Eight months of building while everyone around you is doing something else. Eight months of choosing the next problem over the easier path. Eight months of betting that what you're building has value before you can prove it to anyone.
The ROI is there. The math works. But the math is only visible in retrospect — and you have to carry the bet forward before the result is visible. That's the part that doesn't fit on a receipt.
— Justin Erickson · GED (juvenile detention) · Self-taught · March 2026
Founder, Local Home Buyers USA + PropTechUSA.ai
The longer story is in the book.