Learning is my obsession. My ideal day off is following my curiosity until I stumble into an epiphany.
Having genuine interest is the best way to learn. Unfortunately, “genuine interest” is sporadic and fleeting. That approach leaves you with dynamic overviews of many subjects but without the technical prowess that only comes from overcoming the mundanity of diligent practice.
I’ve been accused of indecisive pursuits and flaky goals abandoned too early. The arms of many loved ones have crossed as I explained how I’d quit yet another project I was passionate about just weeks earlier to pursue something “better.”
But in the last few years, everything has changed — except me.
For an autodidact generalist, AI is a genie in a bottle: limitless questions, instant feedback, no judgments (usually), and an ever-growing repertoire of technical ability.
This flips the script on specialization. Every career dependent on technical expertise is in jeopardy from AI or AI-driven robotics already on the assembly line.
Meanwhile, I’m using AI to fill educational gaps and fuse fields of study. My once-arbitrary curiosities now have the potential to spark new disciplines and industries. See Robert Greene’s Mastery: master two fields, then merge them. Only two?!
When I was studying music in my youth, I’d often skip class for an esoteric book or a meditation too good to interrupt. When high-speed internet arrived, I quit school, became a minimalist hermit, and explored every dim corner of our collective digital consciousness.
When I went back to school for engineering and also enrolled in permaculture courses. When I pivoted to computer science because computers were proving themselves superior at engineering, I started reading business books. When I saw how poorly Comp Sci was being taught, I switched to resource and environmental management, bought a 3D printer, and taught myself to make my own prototypes. One subject was never enough.
When I finally earned a bachelor’s degree, after 10 years’ worth of semesters in higher education, I found many of my uncredentialed “failed” pursuits were perfectly applicable in ways specialists didn’t readily grasp.
I reformulated the standard business plan with permaculture principles, which largely equates to a vertically integrated small business. I aimed to solve a popular environmental materials problem: plastic recycling. This was before AI.
I designed and built devices, programmed automated machines, and streamlined my production process. I analyzed markets, designed products, 3D-printed prototypes, and recorded jingles.
All my arbitrary intellectual pursuits were perfectly applicable. Outsourcing any of those tasks would have been too expensive for my little business, but as a curious generalist, I had a trick up my sleeve the specialist didn’t. And whenever I wasn’t sure what to do, I used my most valuable resource – my ability to learn – to fill the gaps and figure out the best way forward.
Sanctioned Cheating
The guidance counselor misunderstood me. She was recommending help from the Accessibility Services; for people with problems.
When I said, “I’m struggling with the work this semester,” I wanted to express that everyone in my programming class was cheating because the teacher didn’t teach anything and kept yelling “google it” to honest student inquiries. But I was also trying not to be accusatory.
She sounded eager as she rattled off options. “If you have a documented condition, you can have someone read your homework aloud to you, or there are dictation services…”
I blurted, “What?! Disabled people get to cheat?!” She said ADHD might qualify, but my attention had already moved on. “Thanks for your help.”
I wondered how much more productive I’d be if I could read without staring at pages and write without typing. The quality-of-life improvement for my dog and me if I could listen to and draft essays on long walks or while playing fetch would be incredible, I correctly imagined.
That’s when I found the app I’ve used like a foundational Google tool for the last eight years. Despite ElevenLabs and numerous AI voice apps now available (but all paid apparently), I’m still using the modestly named T2S app to “read” books.
I realized my phone already had speech-to-text accessibility options. Immediately began adapting to the practice of listening to assigned literature while in the park and writing essays on log dog walks.
These tools are just as beneficial to the perfectly abled as they are to the tentatively abled.
Intellectually Sustainable Artisan
When I quit school for the last time – aka graduated – my educational drive jumped a magnitude.
I researched, then set up a Precious Plastic-style artisan recycling workshop and got busy making things with my hands while listening to books. On busy workshop days, I’d get through two books: philosophy, marketing, psychology, finance, business, history, how to do this, how to make that.
I was consistently shocked by the indisputable truths absent from my education – details that would have changed the curriculum – and by professors who must have known, but chose not to teach them.
The biggest lesson from my formal education was something I suspected all along but lacked the confidence to claim until I graduated: you can do it without the uni. If you love to learn and you’re entrepreneurial, don’t waste your time and money – especially now that AI is robust.
Get a job where you can listen and learn while you work, so you get paid to learn instead of paying for it.
Given AI’s capabilities, universities increasingly feel like places for people who need structure or accessibility support.
The 2025 Reinvention
In late 2024, I found myself away from my recycling workshop for the first time in four years. Despite all my intellectual side pursuits, I had restricted myself to pondering only during busy-work. There were curiosities I didn’t allow myself to explore for fear I’d abandon my current project in excitement for the next.
On vacation, visiting family, I decided to do what I enjoy most: follow my curiosity. The result was exactly what I feared. Within weeks, I announced I was leaving my last project for a new one – but with good reason. The world was changing and this was my chance to ride the wave.
I learned prompt engineering for definitive outcomes, semantic embeddings, and digital automation. I learned the power of MediaPipe, that it’s available on most phones, and how easy building apps has become. I pushed a Raspberry Pi 5 to its limits while learning about sensors, high-def cameras, and software-optimization tricks.
I learned how to leverage Google’s Edge AI tools for offline, RAG-supported LLMs that run on many phones. I discovered several applications of offline AI that feel immoral not to pursue, like HIPAA-compliant on-device inference for nursing homes.
The only reason I’m not pursuing a project like that is its complexity and funding requirements. So I’m pursuing what I’m calling a fundraising project to finance the others: an augmented-reality marketing project (as of October 2025) inspired by what I learned about WebXR.
After learning that most modern phones support AR and that AR glasses are hitting the market soon, I leveled up my 3D modeling, learned to work with GLB and USDZ files, and leveraged Three.js and Quick Look to build a web app that makes printed media pop off the page via a QR code.
What excites me most about the AR project is getting it mostly automated so I can move on to the next thing: learning advanced robotics.
Takeaways
AI makes specialization a questionable investment. A specialist may be one AI upgrade away from obsolescence – or at least a lot of retraining.
Be a generalist. Make connections. Learn with AI, then listen to the books it recommends while you’re getting paid to do busy-work. Read Anna’s Archive; Meta did.
Now is a time of incredible opportunity if you know just enough to ask the next good prompt and chart your own path. Follow your curiosity before it wanders off. Prompt now.