How long has it been since you last wrote code?

BruceJames — Thu, 10 Sep 2026 08:28:31 +0000

Since 2026, I’ve barely written any code. AI has moved so fast that I haven’t even had time to adjust before it already reshaped the way I work.

Just last year, I was still writing code by hand. I’d think through the architecture on my own, model the business domain, break things down into modules and classes, and weigh the dependencies between them. It was a state of pure focus — every step of reasoning taught me something new, and finishing a design gave me a real, tangible sense of satisfaction from having learned something.

But this year, things have taken a dramatic turn. Much of what I used to do myself can now be handed off to AI. AI can help me sort out the business logic, design the architecture, draw up the UI, write the code, and even get involved in testing and acceptance. It’s almost woven into every step of my product development. What I do most now is audit AI’s output, give it feedback, and correct it where it falls short — especially during implementation, where AI tends to make design mistakes that I have to point out one by one.

Recently I’ve been building a Chrome extension meelang from scratch, with AI involved at every stage — from development and launch all the way to SEO and site operations. When I needed to track user-growth metrics, for example, AI put together a full operations dashboard in under an hour, and once it was hooked up to the database, it was flawless. Building something like that single-handedly used to be nearly impossible. My specialty is programming — I’m no expert in UI design, SEO, or operations — but with AI’s help, all of those problems get solved on the spot, anytime.

What makes AI so powerful is exactly that it helps you excel in areas you’re not good at. It levels the playing field when it comes to access to knowledge. Knowledge itself, no longer an obstacle that holds humanity back.

But the more powerful AI becomes, the more I worry about the future of software developers. It has fundamentally changed the paradigm of development, and maybe soon there won’t even be a “developer” as a profession. Using AI to write code doesn’t actually make you a better engineer — newcomers get no real practice, and the number of experienced developers keeps shrinking. That could easily turn into a vicious cycle for the industry.

In truth, plenty of fields face the same predicament in the face of AI. Designers, animators and filmmakers, actors — even mathematicians, biologists, and chemists. Beside such a powerful AI, we humans seem minuscule. I believe this is a genuine turning point in human history.

I‘m building a Real-Time Translation Tool for Online Meetings. Here's What I Learned.

BruceJames — Fri, 14 Aug 2026 09:34:49 +0000

For the past few months, I've been working on a project that translates online meetings in real time.

At first, I thought the problem would mostly be about choosing the right speech recognition model and finding a good translation API. It turned out those were the easy parts.

The real challenge was making everything feel instant.

People don't wait for subtitles. If the translation appears two or three seconds after someone finishes speaking, the conversation has already moved on. Technically it works, but from a user's perspective, it feels broken.

That completely changed how I approached the project.

Instead of optimizing only for accuracy, I had to optimize the entire pipeline for latency:

Capturing audio continuously
Streaming audio to ASR
Performing speech recognition incrementally
Translating partial sentences
Updating subtitles without flickering
Handling corrections when the speech recognizer revised previous words

Every stage might only take a few hundred milliseconds, but together they determine whether the experience feels "real-time."

Another lesson surprised me even more.

Translation quality isn't just about picking a better LLM.

Spoken language is messy. People interrupt each other, change direction halfway through a sentence, use filler words, and rarely speak in complete grammatical sentences.

A model that performs well on benchmarks can still struggle in a live conversation if the input arrives one fragment at a time.

That forced me to rethink prompts, buffering strategies, and when to display or revise translated text.

Building this project also gave me a new appreciation for streaming systems in general.

Real-time applications are fundamentally different from batch processing. Instead of asking, "How accurate is the result?", you're constantly balancing three competing goals:

Latency
Stability
Accuracy

Improving one often makes another worse.

I'm still learning every day, and there are plenty of problems left to solve.

Over the next few weeks, I'd like to share some of the technical challenges I've encountered, including browser audio capture, streaming ASR, translation pipelines, and reducing end-to-end latency.

If you're building real-time AI applications, I'd love to hear what trade-offs you've encountered as well. Here is the website meelang

DEV Community: BruceJames

How long has it been since you last wrote code?

I‘m building a Real-Time Translation Tool for Online Meetings. Here's What I Learned.