AI-assisted code generation, coding assistants, agentic implementation, and routine bug fixing

AI didn’t replace my development workflow. It changed the way I approach it.

AI coding tools have changed dramatically in the last few years.

Today, developers can generate components, debug errors, explore UI ideas, write documentation, and even make changes directly in a Git repository with the help of AI.

But there is one question I kept asking myself:

How can I actually use AI in my daily projects without becoming dependent on it?

After experimenting with different tools, I eventually developed a workflow that works well for me.

Instead of asking one AI tool to do everything, I use different tools for different stages of the project.

My current workflow looks like this:

Idea → ChatGPT → Stitch → MCP → Antigravity → Jules → Human Review

This article explains how I use each part of that workflow and what I have learned from it.

I'm not a software engineer.

Today I taught a room full of them.

The topic was AI-driven development. Not a demo. The actual practice: how you plan the work when an agent writes the first pass, what you still review line by line, where a human has to stay in the loop, and what changes when writing the code stops being the slow part.

I walked in thinking I was the least qualified person in the room. On raw coding, I was. These folks have shipped more production code than I will ever read.

That ended up being the useful part.

Two things I took from it:

1. AI-driven development is a practice, not a title. It rewards people who think in process, systems, and guardrails. That is my day job. The syntax was never the hard part.

2. Expertise didn't go away, it moved. Nobody in that room got replaced by the tooling. They became the reviewers and the architects, the ones who catch what the model gets confidently wrong.

I left more impressed by them, not less. And proud to have had the seat.
If you're waiting on the right title before you say something about AI at your company, I think you're waiting on the wrong thing.

Curious what you all think. Is AI closing the gap between technical and non-technical roles, or just moving it somewhere else?

#AI #Automation #AIatWork #FutureOfWork #GenerativeAI

AI can write code faster than ever.
But speed is not the same as good engineering.
AI can generate an API, build a component, or suggest a database schema in seconds. What it cannot fully understand is the responsibility behind those decisions.
Will the system still perform under heavy traffic?
Is the code secure?
What happens when a dependency fails?
Can another engineer maintain it six months from now?
Does the solution actually solve the customer’s problem?
These are not only coding questions. They require experience, context, and engineering judgment.
I use AI to explore ideas, remove repetitive work, and deliver faster. But I still review the architecture, challenge assumptions, test edge cases, and take ownership of what reaches production.
The role of a software engineer is not disappearing. It is evolving.
Our value is shifting from how quickly we can write code to how well we can make decisions, manage risks, and build systems people can trust.
AI writes code.
Engineers own the outcome.
How has AI changed the way you approach software development?

AI didn’t replace me — it changed the way I think, build, and solve problems.

A few years ago, my day as a developer looked very different.

I would spend hours searching Google for solutions, reading Stack Overflow threads, watching YouTube tutorials, and trying different approaches until something finally worked.

It was close to midnight on a Thursday, and I had a feature due by 9 a.m. the next morning.

The feature itself wasn't complicated — a discount calculation system for an e-commerce client, applying tiered pricing rules based on order quantity and customer membership level. I'd built things like this before. But the specific combination of rules this client needed was intricate enough that I was still working through the logic well after dinner.

That's when I turned to my AI coding assistant. I described the problem clearly, included the business rules as the client had specified them, and got back a clean, well-commented function that seemed to handle every case. It was exactly what I would have written if I'd had another hour and hadn't been tired. Maybe better.

I believe AI should be forbidden for developers who just push buttons and hope it's fine.

At our company, every engineer uses AI for coding, and we fully support it.

What we don’t allow is blindly shipping code you haven't understood. The moment you stop reviewing what the machine writes, you're not an engineer anymore. You're a button.

Part of the problem is the gap between how the world talks about AI and what it actually does.

For non-technical people, AI genuinely is a breakthrough.

Ask Claude or Gemini for help, and you instantly become a “6” (out of 10) at almost anything. Going from zero to six feels incredible. You type something, software appears, and you think: "I'm a coder now. I can build an app." For non-technical CEOs and entrepreneurs, it's a real dopamine hit.

But for an engineer already operating at a 9, AI writes code they could have written themselves. Sometimes faster, often more convoluted. Useful, not revolutionary. That's why professionals use it as an assistant. Engineers don't mind writing code. Writing code is how they think.

Anyone can prompt "build me a CRM." Maintaining and scaling it is another matter. You stop saying "make the button blue" and start saying "this exact hex code, this spacing, this edge case." Once your English becomes that precise, you've reinvented programming, just in a worse language.

At Satellite, we treat AI like any productivity tool.

If a paid plan makes someone more efficient, we pay for it. We don't track prompts or force anyone to use AI. We measure engineers on output and thinking, nothing else.

Some engineers lean on AI heavily. Some less so. Both are fine. What's not fine is outsourcing your judgment. There's growing evidence that leaning too hard on AI erodes learning, because you're no longer the one solving the problem.

Throw AI at every problem without intention, and you don't get faster engineers. You get lazier ones.

That's the part worth forbidding.

Claude Code vs Codex vs Cursor vs Local AI Coding Agents — my honest experience as a Python developer

AI coding agents are everywhere now.

Every week, some new tool says it can write your code, fix your bugs, understand your project, create full apps, refactor your files, and maybe even replace a junior developer.

But after using AI coding agents in real work, I learned one simple thing:

Most AI coding agents are impressive in demos, but very different when you use them inside your actual project.

A demo project is clean.

Real projects are messy.

Real projects have old files, confusing names, half-written logic, wrong comments, hidden bugs, weird dependencies, and that one file you are afraid to touch because even you don’t fully remember how it works.

So I tested four types of AI coding agents in my own workflow:

Everyone is afraid AI will replace developers.

They're asking the wrong question.

The right question is: which parts of "developer" get replaced, and which parts become more valuable?

AI is very good at writing code when the problem is well-defined.

It is not good at deciding what the problem actually is.

It can give you ten ways to build a feature.

It cannot tell you whether that feature should exist.

What's getting replaced is the part of the job that was always the least interesting anyway: writing boilerplate, looking up syntax, building the same CRUD screen for the hundredth time.

What's becoming more valuable is the part that was always hardest to teach: understanding the actual problem, talking to the people who have it, and making judgment calls with incomplete information.

I use AI every day now. It writes a meaningful share of our code.

But the decisions about what to build, why, and for whom.
That's still entirely human, and I don't see that changing.

The developers who'll struggle aren't the ones who use AI.

They're the ones whose entire value was "I can write this code" and nothing else.

AI didn't make me a 10x developer

It removed hours of boring work.

Now I use it to:
• Design APIs
• Generate boilerplate
• Review code
• Find edge cases
• Write tests
• Create documentation

The biggest productivity gain isn't coding.

It's reducing context switching.

How are you using AI in development?

Let me save you some time upfront: I’m not going to tell you it changed my life.

A month ago, I wasn’t even looking for a new AI coding tool. I already had a workflow that worked. Claude.ai for code reviews, Axios for API calls, VS Code for everything else. Then I kept seeing developers talk about Claude Code — the terminal-based version, the one that reads your entire codebase automatically.

Some said it was the best coding assistant available. Others said it wasn’t worth the price. Most reviews were either suspiciously positive or completely dismissive.

So I stopped reading opinions and started testing.

For 30 days, I used Claude Code on a real production project. Not a side project. Not a tutorial app. A form-heavy medical examination system built for a state government — complex conditional logic, multiple examination form components, Java DTO integrations, file upload flows, and zero tolerance for bugs.