|
|
Monday 6th of July 2026
Software as a Commodity
Eventually, everything turns into soy beans.
Grand reopening. Or so I thought when I started coding this blog. While I was tinkering with my own patches to my code and figuring out the best way to rebuild this damn thing, I had more than enough time to think about the day-to-day grind.
Lately, either I'm wasting too much time or I just can't make time stretch further. I've been working-heavily-on a dynamic inventory system I built for direct sales. Argentina's crisis pushed me back into hardcore programming. Hardcore in the sense that it's built 100% for B2B, and that's where I paused for a second. Competition, even within the ERP space-which I'm building brick by fucking brick-has gotten atrocious. Software has become a commodity. Blame falls on several factors, but it's hard to take the easy route and say, "AI did it." The reason I keep it out of sight in this code-filled world is that I still do my work by hand. But the real factor that made me put AI on the same danger level as weed is that what changed most drastically is customer behavior. Clients aren't the same as before, therefore, neither can the software be. Back then, they'd ask for solutions. Today, they ask for a mix of solution, automation, and ego. You don't fulfill that easily because the request includes "this has to be on my level." Believe me, dear reader, it always is-even if my balls are dragging on the floor after dealing with several clients, it is. You provide a premium service, and in Argentina's case, at an impossible price.
The transformation of code into a goddamn soybean isn't easy to ignore. Sure, freelance platforms and the post-pandemic chaos turned software work into an atomic rarity, and development time and pricing became a nuisance and a burden respectively for many clients. The massive rise of SaaS-software as a service-via subscriptions pushed me into the mess of competing in the same field-and then some, or dinner doesn't hit the table.
This is where I pause, because I'm still working on the second most painful part after the code itself: decent infrastructure.
What I'm building eats up a ton of RAM and CPU due to the nature of asynchronous queries. So, I set up a server at home and threw bandwidth at it to act as both a financial shield and a small central server. From there, I evaluated deploying the software so that people can simply have smart inventory organization and avoid losing money. For the average company, there's no point in busting your brain offering something beyond "I want to make more money" or it won't work. Plain and simple, the average Joe thinks in straight terms like "better TV, better PC, better perks… benefits…" and the list goes on from the common client's benefit to the corporate client's, which is no joke either. In fact, corporate clients expect even better results, which forces you to level expectations by offering a few trial days. "Where did the days go when clients didn't bust your balls to infinity?" I thought that over and over while doing this task, which is, no more and no less, for the only satisfied client I have: myself.
Putting anecdotes and jokes aside: It's true. Clients have become a pain in the ass, and my bottleneck isn't "coding a shitload" but rather, dealing with tech support for under $30 and surviving. I guess it all boils down to an idea I've seen materialized in clients' looks and words more than once: "An AI does this, so it should be cheap."
The general mindset in Latin America doesn't favor the developer. Preconceptions are built around what's an expense and what's an investment, and the pricing scale ends up being determined by a literal imaginary index. No one can speculate more than a guy who doesn't have clear rules demanding his fucking signature on a damn piece of paper. But it's true: software as a commodity forces you to deal with the client's wrong perception of value, even though they already know that for them, value is reflected in money they won't pay accordingly.
My move toward SaaS also responds to the fact that software can no longer be handed over physically, or piracy will eat your liver. That's why we depend on servers and complex structures for something that could run on an office PC.
It wasn't AI. It was the client.
That's right. The client just uses AI as an excuse to tell you "I won't pay that, because it must be easy to make." But it doesn't affect my work, it hasn't eaten it, and I don't think that'll ever happen. Simply put, real demand is in the gutter in the southern hemisphere, and on top of that, speculation is the order of the day. I won't speak for other regions. I'm not there, and I only know a few things from long chats with friends from other parts of the world. That, in my book, isn't enough to give the same diagnosis I'm giving here.
What's happening is that the subscription world distorts things. A lot. The client ends up comparing what you offer to the streaming services they pay for monthly. This is chaotic, but true. Add to that some fierce competition in English, and I just kept developing in Spanish with all the above already in mind.
Still, I'm juggling several projects at once, which forced me to dust off skills from a decade ago and polish them up to do a solid job on servers-which is 70% of the work. The rest is Python, PHP, JavaScript, some framework, and understanding what the client is asking for-when that's even possible.
The commodity pushed me to produce something elaborate. That elaborate thing is, without a doubt, a canned product with an imported vibe, but it's what best fits a new millennium full of potholes in certain practices thanks to an unpredictable and nasty social phenomenon driven by the total dominance of social media. People lose sight of value and focus on aesthetics.
With all that in mind, I took a break to build something of my own while I keep working on other people's stuff. It was a relief.
In just a few lines of code, the PWA-progressive web app-I included in this blog to make it installable on mobile will be able to notify you about new posts. That's good. Also an unavoidable result of expanding my knowledge a ton-inevitable but functional.
This blog will have a lot of my own criticism of what I consider a demolition of value and the involuntary transformation of professional into creator in far too many cases. I'll have more time to talk about that later, but definitely not with the calm of someone putting the final touches on a patent-worthy engine.
Stay tuned, folks, because the content on my blog, though chaotic, might be more interesting than a business chat. Stick around-I'll be back with a less brooding tone to give my opinion on something where being partially destructive in criticism is a good thing: movies.
See you soon!