Modern tooling, the age of rewrites, and where human speed isn't the bottleneck anymore

17 Oct 2026

Years went by since I've posted something in this website, and for a reason: the landscape of software engineering was in a steering internal loop of self improvement on tooling, programming language "wars" and the usual discussion around basic aesthetics of how we typed code, more efficiently or not in our professional lives. Now on 2026, this heavily changed with the introduction of agentic workflows, at the point of many peers reviewing their professional lives and even leaving their careers.

I'm not going to reflect here when or how this all started, but my personal experience was absolutely transformative on mid 2025, when fancy auto complete robots actually started being very useful copy and paste algorithms. I've completely embraced that the "stochastic parrots" had value.

In the meantime, I had a kid, which is now three years old. This taught me a lot and made me realized how precious and important time is for us.

In this new era, the role of the human engineer has shifted from being the primary executor of tasks to being the orchestrator of intelligent systems. The focus is now on designing workflows, setting objectives, and ensuring that the tools at our disposal are used effectively and ethically. Reasons ares plenty, ethical, and understandably the lack of willing of accepting that we are not the ones that are going to dictate and maintain the raw power tools that we used to use in our everyday programmer lives. The pure programmer that used to type 100 words per minute of code has no power as before, the same way as the most efficient and experienced power lifter can't compete on lifting heavy weights against machines designed for that purpose.

One example is: I just dropped Jekyll from this simple website. The goal of having Jekyll was to facilitate the process of generating static pages for blog posts, creating automatic HTML tags that would be semantic, etc. which nowadays, huge models of raw probabilistic computing power can do it in a matter of seconds. Typing isn't the bottleneck anymore. Human speed isn't the bottleneck anymore, as the engineering aspect of building non-physical appliances are now absolutely dedicated of how and when adjust processes instead of manually adjusting tools and tinkering with our own environment.

Anyway, this is basically some couple of random paragraphs and brain dump from me, where I will gladly keep being a software engineer, as I can see it being highly valuable in this new era, where I will continue to do what I've done before: I didn't, and never liked writing code. It was purely brute force. Delivering value and making sense of technical aspects of time consuming tasks that people are not willing to do themselves, due to the lack of interest, time, or expertise on technical details. They don't want to, but I want, which is the entire point of having different human beings being professionals in their respective domains.