From Excel to AI: We Were Always Programmers

Times change. Tools change. But the fundamental question of software development never does: What are we actually agreeing to build?

In the age of generative AI, where the barrier to writing code is collapsing, we need to reframe how we view our craft. The future of software isn't about hoarding syntax; it's about establishing alignment.

1. The Full Cycle of Code: From Syntax to Intent

AI represents the closing of a massive loop. We have transitioned from punching physical cards, to writing low-level assembly, to wrestling with massive framework boilerplates, and now, finally, to simply describing what we want in plain language.

[Punch Cards] ➔ [Assembly/C] ➔ [High-Level Frameworks] ➔ [Intent-Driven AI]

Syntax matters less with every passing year. What matters now is the agreement between stakeholders: users, developers, businesses, and communities.

Code started as a necessity born of military and academic research. It evolved into a highly specialized corporate profession. Now, with AI democratizing creation, it is circling back to being a hobbyist medium once again. And that is incredibly healthy.

• The Corporate Developer ships because they must hit a quarterly profit target.


• The Hobbyist Creator writes because they genuinely care about solving a problem.

Both motives are valid, but when the market crashes and the venture capital dries up, it’s always the hobbyists and the open-source community who keep the web alive.

2. Code as Literacy: Profit vs. Necessity

Corporations are brilliant at scaling ideas. But profit alone cannot be the ultimate metric for the existence of software development. If a tool serves communities, it should be treated as a basic societal necessity—like basic mathematics, or reading.

"We don't debate whether literacy is profitable. We teach people to read and write because it makes everyday life navigable. Code is heading toward that exact same status."

Not everyone will become a professional software architect, just as not everyone who learns to write becomes a published novelist. But almost everyone sends messages, fills out spreadsheets, and runs logical steps in their mind daily.

If you can think logically, you are already executing runtime logic. That is programming.

3. Ownership Is the Real Divide (And Yes, Excel is Programming)

I often meet professionals who sheepishly say, "Oh, I don’t understand code," right before showing me an incredibly complex financial model they built from scratch in Excel.

Here is the secret we don't say out loud enough: Excel is SQL. It is functional programming.

• =VLOOKUP() is a function call.


• A Pivot Table is a database query.


• Cell dependencies are a reactive state graph.
  

=IF(A1="Yes", VLOOKUP(B1, DataRange, 2, FALSE), "Pending")

// This is not "just a spreadsheet." This is a conditional database query.

Technologically underserved communities or countries don't lack engineering talent; they lack ownership of the identity. They have been gatekept into believing that "code" is a mysterious artifact created only by people in Silicon Valley. Meanwhile, they are already building complex logical workflows to run their businesses without the label.

The gap isn't skill. It's the agreement on what counts as "real" development.

4. What Is a Program, Really? (Stop Gatekeeping Localhost)

At its absolute core, a program is nothing more than an agreement on what should be done, translated into automation.

That’s it.

# A cron job running on a Raspberry Pi in your closet:

0 8 * * * curl -s "http://localhost:5000/water-plants"

You can run your own local IoT server on localhost to automate your home. No cloud, no venture capital, no internet connection required. Yet, our industry has trained us to look down on "localhost" projects. We are told that "real" software must be deployed to AWS, scale to millions of concurrent users, and utilize a complex Kubernetes cluster.

We forgot that automation is about simplifying human life, not satisfying cloud providers.

If a single cron job on a $35 Raspberry Pi waters your garden, that is real software. If a shared Google Sheet automatically texts your mother a daily reminder to take her medicine, that is world-class engineering. The cloud didn't make those solutions valid. The fact that they solved a real-world problem did.

5. The New Compromise: Intent Over Syntax

We waste endless energy arguing about "our data," yet we signed it away years ago. Every social media Terms & Conditions screen made it clear: if the service is free, your behavior is the product. That ship has sailed.

The new frontier isn't about hoarding data; it's about establishing a compromise on what we choose to automate, for whom, and why.

AI closes the loop by allowing non-technical domain experts to self-solve. You describe the problem, receive the functional logic, and deploy it—either locally for yourself or globally for your community.

The circle is complete when a non-developer can solve their own operational bottleneck without needing to ask a gatekeeper for permission. That is the moment coding becomes as natural as reading: a basic utility of modern life.

Summary: The Syntax-Free Future

Times change, but the necessity of human compromise and alignment never will. AI is not going to replace software developers; it is going to replace the outdated idea that only developers are allowed to write software.

The future isn't a world where 8 billion people are sitting around debugging memory leaks. The future is a world where everyone can clearly articulate what they want automated, and use whatever tool—be it an LLM, a spreadsheet, or a local script—to make their lives simpler.

Ownership of the problem will always beat ownership of the syntax. Every single time.

What do you think?

Does Excel count as programming in your book? Are we gatekeeping "real" development too much in the age of AI? Let's discuss in the comments below!