Reflection on the future (Spanish version)

Jose Juan MontielJose Juan Montiel

And two more years…​

The map of my curiosity, at 47: Where am I looking now?

A few years ago, I wondered where I would see myself at 50. Today, with 47 springs and 24 years of code behind me, that date is approaching (like when I used to tell myself I’d retire at 30 :) and the answer is not a destination, but a map. Writing helps me pause, refine ideas, and give them an organized vision.

If years ago the focus was blockchain and Golang, my "game board" for the coming months has evolved toward the integration of Artificial Intelligence—but not the kind everyone sees; I want to know what everything is built upon, to understand the "guts".

🤖 The server’s brain: AI and Agents

In recent months, I have set up a personal server that is becoming a cognitive ecosystem. It is not just about executing code, but about orchestrating models:

  • Language Models: Experimenting with Qwen (specifically its Coder version) to boost development.

  • Vision and Multimodality: Integrating VL-Image so the server doesn’t just read data, but "sees".

  • Training Environments: Keeping the Jupyter, Python, and Go stack alive to continue getting my hands dirty with data.

🛠️ Sovereignty and automation tools

Faithful to the "cook it yourself" philosophy, my home infrastructure continues to grow:

  • Privacy and Network: From the essential Pi-hole and H.A (Home Assistant) to managing VPNs and Auto-backups.

  • Content and Networks: Automating the consumption flow with tools for YouTube and a presence on decentralized networks like Mastodon.

🎨 Experimentation: WebRTC, WASM, and Reading

Since I’ve always liked to try new things while learning, I am exploring:

  • Real-time communication: Working with LiveKit and PeerJS to better understand voice and video agents.

  • The future of the Web: I’m still betting on WebAssembly (WASM) and static generators like Hugo.

  • Reading interfaces: Developing more efficient ways to "Read and Listen," integrating "vibe coding" EPUB Reflow with OCR technologies and Whisper/Kokoro for speech synthesis.

Conclusion: The veteran programmer’s Ikigai

As I said in my previous reflections, at 47, my Ikigai remains just that: learning something new every day, sharing it to understand it better, and enjoying the process of building my own digital ecosystem.

🧠 Between Code and Cognition: What makes us human (or AI)?

In my quest to organize my ideas and polish my professional vision, I have been delving into readings that explore the frontier between our minds and the systems I build on my server. If we are going to coexist with AI agents, it is worth understanding what sets us apart.

1. The Art of Remembering (and Forgetting)

In my diagram, Anki and ML systems appear, which reminds me a lot of the memory techniques from "Moonwalking with Einstein".

  • AI and us: While my RAG (Retrieval-Augmented Generation) node seeks information in a cold and precise way, humans need to "specialize" knowledge to retain it.

  • Reflection: Building an external memory system (like my Knowledge Base with OCR and Whisper) is, at its core, an attempt to emulate that "Memory Palace" the ancients used to avoid forgetting what is important.

2. The Learning Algorithm

This also applies to our brain:

  • Learning by teaching: I have found that preparing these notes for the blog forces me to learn better. It is the training of my own mental model.

  • Similarities: Just as I train a Qwen-Coder on my server, these readings suggest that the human being is the most complex optimization algorithm in existence. My goal at 47 is to keep "fine-tuning" my weights every day.

3. The Future: Agents and Consciousness

In the diagram, I have drawn LiveKit Agent and PeerJS nodes. We are moving from passive tools to agents that "listen and speak".

  • The gap: Although my server uses Kokoro for voice and Whisper for listening, the reading makes me question: Where does signal processing end and true understanding begin?

  • My bet: I will continue exploring WASM and Ebiten to visualize these concepts, because as I said years ago: the best way to understand technology is by integrating it into your own professional story.

Summary of Readings: Memory, Cognition, and Intelligence

Group 1: The Theoretical Core (Already read)

These books form the basis of my current knowledge on cognition, memory, and the construction of reality:

  • Thinking, Fast and Slow – Daniel Kahneman: Explains the duality of thought between System 1 (fast and intuitive) and System 2 (slow and analytical), detailing how the brain often makes systematic errors.

  • Moonwalking with Einstein – Joshua Foer: Recounts how to optimize memory through techniques like the "Memory Palace" to transform conscious effort into technical retention.

  • How Emotions Are Made – Lisa Feldman Barrett: Introduces the theory of constructed emotions, arguing that the brain predicts reality and constructs emotions to manage the "body budget".

Group 2: Practical Expansion Vectors (Recommendations)

Suggested books to connect theory with action and mastery:

  • Atomic Habits – James Clear: Works as a manual for programming System 1, turning difficult decisions into effortless automatisms.

  • Peak: Secrets from the New Science of Expertise – Anders Ericsson: Provides the scientific basis and the formula for "deliberate practice" to acquire expert skills.

  • Make It Stick – Brown, Roediger, and McDaniel: Explains that learning must require effort ("desirable difficulty") to be lasting, validating techniques like spaced repetition.

Group 3: Prediction Neuroscience (Close to Barrett)

Books orbiting the idea of the brain as a controlled hallucination machine:

  • Being You – Anil Seth: States that not only emotions, but our entire conscious experience and the "self," are the best guess of a prediction machine.

  • Behave – Robert Sapolsky: An encyclopedic work analyzing the biological context of our actions, from immediate neurobiology to evolution.

Group 4: Additional Mentions

Books used as examples of fundamental structures:

  • Why We Sleep – Matthew Walker: Mentioned as an example of a fundamental book for creating a dense vector space on health and memory.

  • On the Origin of Species – Charles Darwin: Cited as the example of a "Parent Book" that establishes the vocabulary and main axes of a domain (in this case, evolution).

Conclusion (until next time, maybe at 50?)

The point is that I like mixing disciplines, and although the focus of my work is technology, I like to understand the human context, and that leads me to read about psychology, neuroscience, philosophy…​ even carpentry, cooking, or music. I believe that is the key to keeping curiosity alive and continuing to learn every day, regardless of age.

And now without autocomplete, at this age and at this point in my life, and especially with how we have AI right now, I think everything is about to change. It’s not that I think from now on AI will do everything alone, but in the short term, I believe if you don’t jump on the wagon now, another train might not come by. From my point of view, the key is understanding the guts; that’s why I’m starting from the basics, as I’ve done before.

See you soon.