Home
Welcome inside the AI Black Box. It’s brighter in here than you’d think.
Dentro de AI translates to Inside AI, which is exactly what we do here - without drowning in code or calculus.
Hi, I’m a Senior IT Business Analyst who’s spent 30 years deep in tech - and discovered that explaining things simply is the hardest (and most rewarding) part. So I built this space to unpack Generative Artificial Intelligence in plain English - with analogies, visuals, and a dash of humor.
Since launching in early 2025, this site has become a growing library of insights designed for non-IT folks who just want AI to make sense. The star of the site is the How to Learn AI in 3 Weeks course that ties most of the site's content into a Learning Curriculum aimed at non-techies - it mirrors my own learning. Oh, and meet Dentroid - our slightly quirky mascot who helps turn abstract AI ideas into visuals that might or might not fit.
How to Learn AI in 3 Weeks
This is the complete Learning Curriculum that onboards you within only 3 weeks to the basic AI fundamentals:
- What is happening inside the hidden AI Black Box?
- How can I use that knowledge to better interact with AI models?
- How come that AI is sometimes genius, but other times dumb? And why?
- Token, inference, context - what is that all about?
See the course structure here:
Most of the content of this site has been waved into this coherent curriculum - it introduces the Mental Model: BEFORE → INSIDE → AFTER the AI Black Box, which assists nicely to explain how AI models work. Jump right to the course or read how it came about.
You find all site content via the navigation, but let me lay it out here too:
Blog Articles:
How AI Large Language Models (LLM) Work
- What happens when you prompt ChatGPT?
- Explained in plain non-geek English.
AI Model Lifecycle - from Creation to Inference - Explained with Car Analody
- The 9 steps of LLM lifecycle - from architecture design to user inference.
- Explained with an analogy from the industry: Automotive Manufacturing.
Understanding Open Source in AI Models
- Many models claim to be Open Source, while in reality they are just Open Weights - and why that is not enough.
The Big AI Players - the Who Is Who of the Industry
- Who forms the industrial AI landscape?
- Introduction and background - from public companies to research institutes.
AI Visualizations
- Interactive Websites and Tools - plus Videos to explain AI.
- From single Token to high level architecture of deep learning systems.
AI Timeline
- How did we get here? Including entries you would not neccessarily expect.
- All big moments and events - from Research Papers to Pop Culture.
AI Glossary
- AI/ML has a very special vocabulary - this is the place to get familiar with it.
- With Hundreds of Foundational Terms and their plain-English Explanation ... plus a Quiz!
AI News
- A curated roundup of AI’s fast lane - crafted with a model’s help.
- Spotlighting the biggest launches, research milestones, and what’s making waves in the industry.
About
- Background on Dentro de AI Project and Author.
Future Content Ideas
Now that the Learning Curriculum is complete I'm aiming to dive deeper:
- AI Installation Use Cases for real world Business Solutions.
- Explainable AI / Interpretable AI / Mechanistic Interpretability
Easy to Use, Hard to Understand?
Artificial intelligence can feel deceptively easy. With just a prompt or a few lines of code you can generate images, analyze data, or create text that feels surprisingly human. But beneath that simplicity lies a world of complexity: statistical models, neural networks, training data, biases, and massive computing power. This contrast is part of what makes AI both exciting and overwhelming. The goal of my Dentro de AI project is to bridge that gap - to demystify the underlying concepts without losing sight of the power and nuance behind them.
AI Black Box de-mystified
For AI to be truly useful and widely adopted, people need to understand it - not just developers and data scientists, but business leaders, creatives, and everyday users. What is happening inside the Black Box? Easy-to-grasp explanations build trust, reduce fear, and open the door for Learning Curriculums or even Career Paths. This project is built on the idea that easy access to knowledge is the foundation for meaningful AI usage.
Remember the first Office PCs? The Dot-Com Era?
If you've been in tech for a while, today's AI boom might feel familiar. It's like when the personal computers entered the corporate world - geeks first, then business, often without a clear strategy. The Internet followed a similar path: early adopters made noise, businesses scrambled to catch up, and real value only emerged when understanding met purpose.
AI seems to be going through the same cycle. Tools are available, expectations are high, and many organizations are still figuring out what it all really means.
The technology behind AI is complex, fast-moving, and often hidden behind marketing terms or black-box models (closed source). But to make smart decisions, you don't need to become a machine learning engineer. You just need a clear, working understanding of what's happening under the hood - concepts like models, training data, tokens, embeddings, and inference. Knowing these basics helps you to manage expectations, excel at prompting, spot opportunities, and be able to make more informed business decisions.
This space is here to help with that - not by simplifying things to the point of distortion, but by breaking them down into something you can actually follow. And by keeping pace as the field evolves.
Breaking Down the Jargon
If you're just getting started, the AI Glossary is a good first step to scroll through. It covers core terms and ideas - explained simply, without dumbing them down. Whether you're trying to understand what a transformer does, what fine-tuning means, or why everyone is suddenly talking about tokens, you'll find practical, business-relevant explanations including links for further study.
Pro Tip
Type F anytime on this site to activate the Search Box in case you want to look up an AI/ML term. 
Powered by: mkdocs-material
Some words on what I use for here publishing: when researching for an easy but powerful tool I decided on mkdocs-material - for its rich features, speed and easy maintenance. I'm a huge fan of markdown and Python - and mcdocs is built on both. Back in the 90s you had to run your own Unix to locally test, develop and deploy webpages - these days it comes all bundled and ready-to-go. If you are looking for a blogging or documentation tool too: check out the experimental mkdocs-material Features / Showcase page I did for demonstration purpose.
Let's Connect!
Have a term you'd like to see added? Spot something unclear? Have an idea to collaborate? Let me know! An additional purposes of this site is for me to network and collaborate - so please don't hesitate to reach out:
I'd love to hear from you!
