I’ve been working with React since it came out. Back then, frontend meant components, props, and making things look right. We were happy. We were innocent.
Then one day someone said, casually, over a beer:
“Frontend became a wildcard.”
My brain did the Windows XP shutdown sound.
When I started programming, he was Avatar Aang. I was Zuko — writing questionable JavaScript, shipping bugs like it was my job. Classic junior dev energy. Almost ten years later, he drops that line like it’s obvious.
Frontend Was Never Just UI
At some point, frontend stopped being about pixels and started being about decisions.
What data do we show? When do we fetch it? What happens when it fails? How fast does this feel to a real human?
Those aren’t backend questions, and they’re not design questions either. They’re product questions, answered in code.
Backend still does the heavy lifting: data, rules, guarantees. Frontend is where all of that shows up for the user. This blog has always been about how those pieces fit together, not about picking sides.
Frontend lives where users, systems, and business goals collide. That forces you to think in flows, not screens. In outcomes, not components.
Enter AI (The Tony Stark Moment)
AI didn’t replace my frontend skills. It mostly removed friction.
I could prototype faster, refactor with more confidence, catch edge cases earlier, and think about architecture before writing code — including realizing earlier when something probably belongs in the backend.
It felt less like autocomplete and more like having someone to think out loud with. That’s when frontend stopped being reactive and became more intentional.
From Components to Flows
At some point, I stopped thinking in pages and started thinking in flows.
Authentication flows, loading states, error recovery, user trust.
Knowing some backend made me better at frontend. Thinking carefully about frontend made my backend decisions cleaner. AI didn’t erase the line between them, but it did make you notice it more.
Frontend Is a Perspective Now
Frontend today isn’t about knowing a framework. It’s about understanding how decisions ripple through a system — database, API, network, UI — and how they feel to the person using it.
AI didn’t kill frontend. It just changed how you think about it. And once that clicks, it’s hard to unsee.
The Next Level: Going Deeper Into Machine Learning
Eventually, curiosity kicks in.
If frontend is a perspective and AI is an amplifier, the next step isn’t just using models. It’s understanding them enough to know what they’re actually doing.
That’s where machine learning comes in. Not in a PhD sense, but in a very practical, “what problem am I trying to solve?” way.
I trained and deployed a small sentiment analysis model using DistilBERT to analyze Google Maps reviews and see how people express opinions in the wild.
Here’s the experiment:
https://huggingface.co/spaces/wizsebastian/distilbert-sentiment-luis?logs=container
Is it perfect? No. Is it useful? Maybe. Is it interesting? Definitely.
That’s usually a good place to be.
Top comments (0)