Facundo Uferer
I never believed that artificial intelligence was going to replace developers outright. My view has always been slightly different: AI will replace, or at least leave far behind, those who refuse to learn how to use it. Not because a machine will wake up tomorrow morning eager to take our desk, but because an engineer who knows how to leverage these tools can research, design, implement, review, and learn at a velocity that is impossible to ignore.
In that context, I closely followed the work of Alan Buscaglia, creator of Gentleman Programming. What I found was far more valuable than yet another evangelist hyping that AI will change everything: a genuine teacher for those of us who want to learn how to work with it effectively.
His ecosystem can be explored at Gentleman Programming. What stands out about his contribution is that it never gets stuck in prompt tips or flashy demos. Alan advocates integrating AI into the engineering process with discipline, structure, and concrete tooling. The goal is not to let an agent blindly write code while we passively watch our profession erode. The goal is learning how to direct it.
Beyond Autocomplete: Gentle-AI and Engram
That vision comes to life with Gentle-AI. For me, it was a turning point in viewing AI as an authentic development suite rather than an isolated chat window. It orchestrates agents, skills, persistent memory, task planning, and project workflows. The experience feels much closer to having a personal Jarvis than dealing with an overenthusiastic autocomplete.
The other pillar is Engram. Its mission is providing persistent memory to AI agents, retaining architecture decisions, context, trade-offs, and discoveries across sessions. That memory fundamentally shifts how we interact with the tool: it is no longer about one-off prompts, but about building compounding continuity. In real-world projects, remembering why a design decision was made three weeks ago is infinitely more important than spitting out boilerplate in five seconds.
The Commitment to Open Source
There is another element I deeply respect: both projects are rooted in free and open-source software. The insights are not locked behind paywalls, private demos, or closed courses. The source code is out in the open to study, experiment with, tweak, and contribute to. In an AI ecosystem increasingly dominated by walled gardens, open foundations matter.
Alan brings something far deeper than utility scripts: a coherent philosophy for coexisting professionally with AI. Not driven by fear of obsolescence, nor by naive illusions that AI will solve every hard problem for us, but anchored in pragmatic reality: engineers who master these tools will hold an undeniable advantage over those who choose to ignore them.
Keeping an Open Mind and Experimenting
My recommendation to any developer is simple: stay curious and experiment. Visit gentlemanprogramming.com, spin up Gentle-AI, install Engram, inspect the code, and build your own workflows. You don’t need to swallow every breathless prophecy about AI. In fact, healthy skepticism is healthier. But you definitely need to learn how to direct it.
Perhaps the real question was never whether machines will take our jobs. The truly uncomfortable question is: what happens when the engineer next to you learns to wield their own Jarvis while you keep coding as if Jarvis never existed?
For anyone who doesn’t want to find out the hard way, Alan Buscaglia is one of the best mentors to start with.