menu
I No Longer See Software as “Just Software”

I No Longer See Software as “Just Software”

posted 1 min read

How research and green software engineering changed the way I see code

For a long time, I thought software engineering was only about features, performance, and shipping products.

Today, I realize it’s much deeper than that.

A small shift in perspective

I used to measure good software by how fast it runs, how clean the UI looks, and how many users it serves.

But after working on topics like GNSS-based traffic modeling and green software engineering, something changed.

I started asking different questions:

• How much energy does this system consume?

• What invisible resources are being wasted?

• Who is affected by the decisions we make in code?

That’s when software stopped being “just software” to me.

Software writes real-world consequences

When you model traffic data, you’re not just optimizing routes — you’re influencing pollution, time, safety, and city behavior.

When you analyze software energy consumption, you’re not just profiling code — you’re touching sustainability, infrastructure load, and environmental impact.

Code doesn’t stay on the screen.
It moves into streets, servers, electricity grids, and human lives.

⚙️ Engineering is no longer neutral

Every design choice has a cost.

• One inefficient loop multiplied by millions of users

• One always-on service running 24/7

• One poorly planned system scaled too fast

Software engineers are now part of global systems, not isolated applications.

And whether we notice it or not, we are shaping:

• energy consumption

• digital behavior

• economic decisions

• and even environmental outcomes

What this changed for me

Today, when I write or analyze software, I don’t only think about:

✔️ “Does it work?”

✔️ “Is it fast?”

I also think about:

✔️ “Is it responsible?”

✔️ “Is it efficient beyond performance?”

✔️ “Is it solving the right problem?”

Final thought

Software engineering is no longer just about building products.

It’s about designing systems that live inside the real world.

And once you truly see that —you can’t unsee it.

What questions do you have about responsible software? Drop a thought below — I’d love to hear your take.

You can find me on GitHub: https://github.com/fadydesoky

4 Comments

•

That line about code moving into streets and grids really stuck with me, Fady Desoky Saeed Abdelaziz. Makes you pause before calling something a small design choice.

2 votes
•

@[Ben Miller]
Thank you, Ben — I really appreciate that.

That was exactly the idea behind that line. A lot of what we call “small” decisions in software quietly scale into real-world systems.

I think once you connect code to its real-world footprint, it becomes harder to treat design choices as purely technical anymore.

0
•

For me, the key is making sure we’re solving the right problem before deploying any software. I really appreciate this perspective —

I like this post.

1 vote
•

@[edmundsparrow]
I completely agree.

Solving the wrong problem efficiently can sometimes do more harm than good.

Responsibility starts long before code — at how we frame the problem itself.

Out of curiosity, what helps you decide whether a problem is worth building software for?

1
•

I usually think of it like lifting weight.

Is the problem actually heavy enough to need a machine, or can a person handle it just fine?
And if it is heavy, does it need to be lifted over and over again, or just set down once?

At Gnoke, we lean toward lightweight solutions — ease of use, clarity, and freedom over complexity. If software doesn’t make things simpler or freer, it’s probably the wrong tool. My honest opinion.

1
•

@[edmundsparrow]

The “lifting weight” analogy is really powerful — especially the point about frequency.

I also strongly agree on lightweight solutions. Most problems today aren’t about what software can’t do, but about how heavy and intrusive it quietly becomes.

Really appreciate this perspective, edmundsparrow.

1
•

Great one, Fady!
For the past few days, I have been thinking about how much electricity and computing resources systems like ChatGPT consume at scale. After reading your post, it makes me question whether the cost of running AI is truly worth it. Maybe it is, or maybe it isn’t. I guess only time will tell.

1 vote
•

@[Vignesh J]

That’s a really important question, and I’m glad you brought AI into the discussion, Vignesh.

I think the real issue isn’t whether AI is “worth it” in general, but what kind of AI is worth it.

There are use cases where the social or scientific value might justify the energy cost (healthcare, accessibility, education, climate modeling, etc.).
And there are other cases where we’re clearly over-using heavy systems for problems that don’t truly need them.

For me, responsible software starts with being honest about that trade-off:
not just “can we build it?” but “should this problem consume these resources?”

Maybe the future isn’t less AI — but more efficient, more intentional AI.

I’m curious: what kinds of applications do you personally feel justify that cost?

1
•

@[Fady-Desoky-Saeed-Abdelaziz] In my opinion, I agree with you.
Areas like healthcare and education definitely justify the cost.
But using AI to replace developers, or leaning too heavily on prompting for routine problems, doesn’t seem like a responsible use of resources to me.

1
•

@[Vignesh J]

Well said — I completely agree, Vignesh.

It’s not about “AI everywhere,” but AI where the value truly justifies the cost.

Appreciate this perspective.

1
•

Using a voice to text or sending a voice note?
Either way you just record your voice.

When is is necessary to use either of the two?

Are the two supposed to exist or only one?

1 vote
•

@[edmundsparrow]

And that’s actually a really good concrete example, I think cases like voice notes vs voice-to-text show exactly what you described earlier: sometimes we build more system than the problem really needs.

For me, the responsible question isn’t “can both exist?” but “what real friction are we removing, and at what cost?”
If the goal is just capturing a thought, a lightweight voice note might already solve it.

Adding transcription, storage, models, and pipelines should only happen when it clearly unlocks new value — not by default.

Otherwise, we slowly turn simple human actions into heavy infrastructure.

1

More Posts

Green Software Engineering: Measuring Energy, Not Just Performance

Fady-Desoky-Saeed-Abdelaziz - Jan 7

My Research Journey in Green Software and Urban Mobility — Fady Desoky Saeed Abdelaziz

Fady-Desoky-Saeed-Abdelaziz - Jan 8

Before optimizing code, optimize the process

Fady-Desoky-Saeed-Abdelaziz - Jan 14

Green Software in Action: Building a Universal, Driverless OBD-II Interface

edmundsparrow - Jan 14

Mistakes I made in my Software Engineering career

rumaisanaveed - Jan 23, 2025
chevron_left