Turbocharging self-learning with AI
jenxi.com Vision alchemist crafting strategic innovation & AI adoption. Bridging startups to China’s ecosystem advantage. Building a cyberbrain. Registered pharmacist × Brand strategist × Storyteller
There was a time, not so long ago, when figuring out why your JavaScript code produced that peculiar “undefined is not a function” error involved a mystical journey through the hallowed shelves of your local library.
Picture it: a harried web developer, eye-strain headache fresh behind the forehead, thumbing through a dog-eared copy of “JavaScript: The Definitive Guide, 1996 Edition” if they were lucky enough to avoid having to wait line for someone else to finish it. After finding a potential solution, they proceed to jot it down in their trusty little notebook as they prayed that the solution would work when they got back on the computer. It wasn’t efficient, but by God, it had character.
Fast forward to 2025, and our intrepid developer is now casually chatting with an AI about why that same error has evolved into a more concerning “Cannot read property ‘map’ of undefined.” The AI doesn’t judge their curious questions, it simply offers three possible solutions, complete with code snippets and a friendly reminder about that pesky asynchronous data fetch they keep forgetting to await.
The Library Era: When knowledge smelled musty
In the late ’90s, when I was teaching myself web design with all the grace of a giraffe on roller skates, learning a new skill meant one thing: becoming intimately familiar with your local librarian, or in my case becoming one. I volunteered as a librarian at the school library and later on the public library near my school after school hours so I could get my hands on the latest acquisitions and to intercept popular books that were always on loan.
For web designers, it was the sacred For Dummies series and O’Reilley’s animal-themed programming books (because nothing says “JavaScript fundamentals” like a bewildered rhinoceros on the cover). And Photoshop 5 Bible (Photoshop 5 not CS5, they are published 12 years apart), for anyone who was desperate like me to add more animated GIFs to their Geocities shrine.
Trying to visualise solutions for programming and graphic design, and then attempting to recreate them when you got back on the cousin’s or neighbour’s family computer. Or attempting to make crude notes of what the images in the books depicted (imagine if you had a nifty pocketable device that could capture whatever you saw so you could refer back to the photos anytime), or attempting to photocopy more than the allowed number of pages for reference books, while the librarian was trying to help some wailing toddler find their parents.
These are art forms forever lost in the modern world.
The Digital Enlightenment: Websites and forums
As the new millennium dawned—and after we collectively survived the anticlimax of Y2K by staying up past midnight—the learning landscape shifted dramatically. Suddenly, knowledge wasn’t bound in leather and paper but floated freely in the digital ether, waiting to be plucked by anyone with a 56k modem and the patience of a saint.
For web designers, this era birthed legendary sites like A List Apart, Webmonkey, and CSS-Tricks, where you’d spend hours trying to understand why your div was floating right off the page like a helium balloon cut loose at a child’s birthday party. Forums spawned digital communities where username “HTMLWizard99” might be a 12-year-old prodigy or a 30-year veteran developer—you’d never know until you tried their oddly specific advice to “add an empty div with clear:both and sacrifice a small goat to the Box Model gods.” It was a glorious, chaotic time.
I spent a big chunk of my free time trawling websites like Photoshop Cafe, Good-Tutorials, Abuzeedo, PSDTuts, and forums on DeviantArt, CGTalk, and Renderosity to find the latest tips and tricks to create skeuomorphic designs.
You only had text and accompanying static images that tried to explain how you used certain tools, with the rare GIFs that was a luxury both in the bandwidth cost and time needed to wait for it to load. Don’t get the results? Scroll back to the top and start from the beginning. After a few failed tries, you leave a comment and waited until the author halfway around the globe woke up and saw your question. If you’re lucky, someone who stayed closer to your timezone replied that pointed you down the right direction.
After time, with the accumulation of knowledge, sites like StackOverflow, GitHub, and Reddit emerged as places to search for existing solutions. Someone had the same problem as the one staring at you, and different people have shared how to fix the problem. You could just copy and paste the solution instead of having to wait for replies.
The YouTube University: Visual learning revolution
Around 2010, a new phenomenon emerged: suddenly, everyone was learning not by reading, but by watching someone else do the thing first. YouTube became the world’s largest technical college, with a faculty of enthusiastic amateurs and grizzled professionals alike. Educational video content then spilled over to TikTok and Instagram Reels later in the decade.
For web developers, this was revolutionary. No longer did you have to interpret cryptic documentation like “implement a responsive flex container with appropriate breakpoints”—now you could watch the screencast while explaining exactly which CSS properties to use and why Internet Explorer would inevitably break it anyway.
Learning JavaScript frameworks became less like deciphering ancient hieroglyphics and more like cooking along with a slightly over-caffeinated chef who built a full e-commerce site in 20 minutes.
You still had to wait for replies to your comments, but you get to pick up certain habits and workflow while watching someone else work. I’ve lost count of the number of Photoshop and Figma shortcuts and hotkey I’ve picked up from YouTube videos. Similar to how I learnt strategy and tricks for Red Alert 2 while hovering my cousin or friend’s shoulder while they took out the Tesla Coil with Tanya.
But more often than not, the top upvoted comment is probably the question you were going to ask anway.
The AI Companion: Learning through conversation
Which brings us to our present moment of educational evolution. The AI boom has become a developer’s best friend, like having an infinitely patient mentor who never tells embarrassing stories about that time you forgot a semicolon deployed code that took down the entire website.
Today’s learning process has been condensed from “spend hours searching for an answer” to “ask the AI to write three possible solutions, modify your code, and get it to double check its work for errors.” It’s like having a senior developer, a technical writer, and a cautious project manager all rolled into one digital package.
For web developers specifically, AI has become the ultimate coding companion. It can check through your code for errors, suggest ways to improve and streamline the code, and even help you brainstorm future features to add to your roadmap. All without making a single joke about your choice in naming variables (looking at you, finalFinalActualLastVersion2).
Beyond just finding your answers
The AI advantages extend beyond just information retrieval:
- Contextual understanding: Unlike a static documentation page, AI can follow a conversation about your specific problem, asking clarifying questions about that weird bug that only appears when a user is logged in, using Safari, during a full moon.
- Code analysis: Many AI systems can now analyze your spaghetti code that would make Italian chefs weep, identifying anti-patterns and potential performance issues that might look like perfectly reasonable solutions to your sleep-deprived brain.
- Customized learning paths: Rather than following a one-size-fits-all tutorial, AI can adapt its explanations based on your experience level, explaining what a variable is to beginners or jumping straight to advanced topics like monads and higher-order components for the masochistically inclined.
- Multilingual support: For international developers or those working in diverse teams, AI translates technical jargon instantly, making knowledge accessible regardless of whether you learned to code in English, Mandarin, or the ancient and mysterious language of PHP.
To me, AI tools like Cursor and Windsurf are more than just your co-workers and interns who help you grind out the hard work while you supervise and direct the overall project progression. They can be your business partner or coach by helping you walk through your thoughts, suggest ideas, question you to test your resolve and make you rethink your approach and direction.
The Rise of “Vibe Coding”
This AI revolution has even sparked entirely new approaches to programming itself. Recently, AI researcher Andrej Karpathy coined the term “vibe coding” to describe a workflow where developers essentially surrender to the AI’s suggestions, focusing on outcomes rather than scrutinizing every line of code.
It’s a fascinating shift in how we think about programming—one where intent matters more than implementation details. Instead of meticulously crafting each function, developers can focus on high-level design while the AI handles the mechanical aspects of code generation.
I’ve only touched the surface of this phenomenon here, but I’m exploring vibe coding’s implications, benefits, and potential pitfalls in a dedicated article on vibe coding. Whether you see it as programming liberation or concerning shortcut, there’s no denying it represents yet another transformation in our relationship with technology and learning.
Learning evolution: From book stacks to binary stacks
The journey from library bookshelves to AI chatbots represents more than just a change in medium—it’s a fundamental shift in how we approach learning. We’ve moved from passive consumption of knowledge to interactive problem-solving, from generalized information to personalized guidance.
For the red-eyed heroes of the coding world, this transformation has been particularly profound. The developer who once had to memorize API documentation or lug around a 1,200-page “Definitive Guide” can now access the collective wisdom of the programming community through a device that fits in their pocket. It’s like having every experienced developer who’s ever lived standing beside you at the keyboard, whispering suggestions and gently steering you away from catastrophic Git pushes directly to main.
Beyond answers: Understanding principles
The real magic, though, isn’t just in having answers—it’s in the learning process itself. When you ask an AI why a particular solution works, it doesn’t roll its eyes and just tell you “because that’s how it’s done.” It explains the underlying principles, the cause-and-effect relationships, the computer science concepts that make modern web applications both marvels of human ingenuity and absolute nightmares to debug.
It’s like having your computer science professor, your senior developer mentor, and that one brilliant but socially awkward colleague all merged into one entity that never gets tired of your questions.
The digital apprenticeship model
In a way, we’ve come full circle. The apprenticeship model that built craftsmen for millennia—learning directly from a mentor through conversation, observation, and guided practice—has been reborn for the modern age. Only now, your patient mentor never retires, never has a bad day aside from the occasional hallucinations, and doesn’t make you refactor legacy code before imparting wisdom.
For today’s aspiring developers, designers, and self-taught enthusiasts of all stripes, the AI boom isn’t replacing the joy of learning—it’s supercharging it. Gone are the days of learning in isolation, piecing together fragmented knowledge from Stack Overflow answers that begin with “This question has already been asked…” Today’s learners are having conversations, solving real problems, and building understanding through interactive guidance.
A new generation of autodidacts
As I look back on my journey from library-dwelling novice to AI-augmented creator, I can’t help but feel a twinge of envy for those just starting their learning adventures. They’ll never know the unique frustration of spending three hours hunting down a missing closing tag, but they’ll experience something perhaps even more valuable: the freedom to focus on understanding rather than merely finding information.
The toolbox has changed, but the essence remains the same: human creativity, now with a digital mentor-intern that never sleeps, never forgets, and thankfully, never commits directly to the production branch without a code review.