The future of software is written in plain English
I used to think my bottleneck was coding speed.
Now I think it was clarity.
Last month I built a handful of free tools for Nimbflow, my agency, in a few hours. Not by writing them line by line. I typed what I wanted into a chat box. The AI built it. I read through the code, fixed a couple of things, and shipped it.
That's when it hit me.
English has quietly become the most useful programming language I know.
I'm not the only one noticing it either. AI tools are turning plain language into the interface for building apps, automating workflows, querying databases, even writing infrastructure.
This isn't no-code. It isn't low-code.
It's zero syntax, all intent.
This isn't a hypothetical anymore
Function calling lets you define what an API does, then have the model call it with the right parameters from a plain-text question. You don't write get_weather(“Singapore”). You just ask, “What's the weather like in Singapore right now?”
The model does the translation. It knows which function to call and fills in the blanks.
String by Pipedream goes further. You can type this:
“When a new row is added in Google Sheets, send me a summary in Slack and create a draft email.”
It builds the whole workflow. You're not wiring inputs to outputs anymore. You're describing what you want to happen.
Lovable builds full-stack apps. You write:
“I want a basic CRM with a signup form and a notes field per contact.”
It scaffolds the project, writes the backend, styles the frontend, and deploys it. You get a URL.
No sandbox. No boilerplate. Just a working thing.
People are building entire products this way now, from a chat window.
One person doing the work of ten
This is the part that gets me.
I'm not trying to become a better developer. I'm trying to move faster with less.
Natural language coding does that.
You go from idea to prototype in hours instead of weeks. No picking a framework. No spinning up a backend. No hiring a contractor.
You describe what you want. You check what comes back. You tweak it. You launch.
For me, this is the real shift. It changes what one person can actually run alone.
I can test more ideas. I can customize things per client without opening a dev ticket. I can handle ops, onboarding, and content without a queue of tasks sitting on someone else's calendar.
I don't need to be technical. I need to be clear.
Clarity is doing the work that a computer science degree used to do.
Where this still falls apart
Natural language is powerful. It's not reliable.
Not yet.
AI agents still hallucinate. They write code that looks right and fails quietly. They misread ambiguous instructions. They'll run a bad plan with total confidence and no warning at all.
So my job isn't gone. It moved.
I'm not writing code anymore. I'm reviewing it. I'm prompting clearly, then debugging the AI's logic instead of my own.
Some problems are still too complicated to hand off this way. Anything that needs real performance tuning, low-level access, or logic that isn't obvious, you still need an actual developer for that. This covers most of what I build. Not all of it.
And a product you built with one sentence isn't safe just because you built it fast. If you can build it with a prompt, so can whoever copies you.
Execution still has to be earned by hand. So does the actual experience of using the thing. So does distribution. AI gets you to a working version one. Getting past that still takes real thinking.
The interface changed. The code didn't.
Nothing here replaces engineers.
What's shrinking is the distance between having an idea and having it built.
Natural language is the interface now. It's how you tell the system what to do. Code still runs underneath all of it. You're just not the one typing it. You're directing it.
The people doing well with this aren't the best coders. They're the ones who can describe a system clearly enough that a model can build it, and who notice when the model got it wrong.
Your job becomes explaining a system to something that can build it. The clearer you are, the better what comes back.
That's just being able to think in systems.
I stopped worrying about my stack
If you're still fighting with syntax for basic workflows in 2025, you're behind.
People assume that's lazy. It just makes me faster.
And right now, speed matters more than polish.
So I've stopped obsessing over tech stacks. I spend that energy on thinking clearly and saying exactly what I mean.
Because the future of software isn't being written in Python, JavaScript, or C++.
It's being spoken. In English.
And it already started, whether I was paying attention or not.