Research & Analysis
Faster immersion into complex topics, technologies, and solution approaches — with critical evaluation.
Honest About AI
Yes, I use AI
AI is one of the most significant technological developments of the past few years. Nearly everyone in tech is using AI in some form today — and that makes sense.
The important distinction isn't whether someone uses AI or not. The important distinction is whether AI is used professionally, with control and accountability — or uncritically, without review, and without understanding the output.
This page explains how I use AI in my work. Openly and concretely.
Casual vs. Professional
Not all AI use is equal
You may have heard the term "vibe coding": describing what you want to an AI tool in rough terms and letting it generate the code — often without truly reviewing or understanding the output.
Where it can be useful
- Fast experimentation and prototyping
- Lower barrier to entry for new technologies
- Creative impulse and fresh momentum
Where it becomes a problem
- Loss of control over the codebase
- Hidden bugs, security vulnerabilities, and fragile logic
- Code that "works" — but isn't maintainable, secure, or scalable
- Lack of understanding of what is actually happening
Vibe coding can be useful for experiments and early prototypes. For production software that people rely on, it's not enough. That's where the real work begins.
How I Use AI
AI as a teammate — not a replacement
AI doesn't replace competence — it amplifies it. I use AI like a skilled colleague: as a sparring partner, research assistant, and accelerator. But every suggestion is questioned, every output is reviewed.
Concept & Ideation
Sparring partner for architecture decisions, brainstorming, and strategic questions.
Code Assistance
Help with debugging, routine tasks, and boilerplate — always under manual control and review.
Documentation & Communication
Clearer wording, better structure, and more accessible technical writing.
The Human Steers
Control, Correction, Accountability
Professional AI use doesn't mean writing prompts and accepting the output. It means: understanding what the AI does — and stepping in when needed.
A modern aircraft has autopilot. But no responsible pilot relies on it blindly. Autopilot assists — but the pilot must always be ready to take manual control. That's exactly how I work with AI.
Critical Review
Every AI output is read, understood, and evaluated — before it enters a project.
Manual Correction
Weak logic, poor structure, and hidden errors are caught and fixed by hand — or rewritten entirely.
Deep Understanding
The real skill isn't prompting — it's evaluating, guiding, and taking over manually when needed.
Quality Assurance
Automated tests, code reviews, and human judgment form a real quality safety net.
Clear Boundaries
What AI is not allowed to do
No Blind Copy-Pasting
AI-generated code is read, understood, tested — and rewritten when necessary.
No Data Sharing
Confidential client data is never submitted to external AI services.
No AI Autopilot
No output goes into production unchecked — no matter how convincing it looks.
Why This Page Exists
Transparency isn't optional
Many companies and freelancers use AI — but few talk about it openly. This page exists on purpose.
Honesty
Honest clients deserve to know how their projects are actually built.
No Hiding
AI use is neither concealed here nor sold as magic. It's a tool — nothing more, nothing less.
Trust Through Openness
The important question isn't whether AI is used — but whether it's used wisely, responsibly, and under professional control.