Works in the demo, costs in production
The prototype is impressive. Tokens are invisible. Errors are rare. Then production starts, and the bill changes character. Klarna learned this the hard way in 2025.
Artificial intelligence, taken apart.
Between what you're sold and what it does.
This site sells nothing. It takes things apart. Each article starts from a verified fact and goes where the commercial pitch stops. A project by Privateer.
Latest
Where AI is useless, and where the market says otherwise
Often, a regex, a business rule, or a sort is enough. AI is sometimes sold as a universal answer to problems with simpler, more reliable, cheaper solutions. A tour of cases where the market sells you what you don't need.
Build / Buy / API: the trade-off nobody frames correctly
Building your own model, buying a vertical solution, or using an API: three strategies with radically different risk, cost, and sovereignty profiles. The sovereignty angle often changes the decision.
The 3 questions to ask before adding AI anywhere
Not every AI deployment decision deserves a 6-month project. Three questions filter relevant use cases from marketing use cases in under 30 minutes.
The real cost of an AI project: what the quote doesn't show
The AI integration quote only shows the surface. Underneath: token costs in production, infrastructure, data preparation, prompt maintenance, human review, and technical debt. The real TCO is often 3 to 5 times the initial budget.
The AI pitch jungle: reading it without being fooled
Every software vendor now has an 'AI' offering. Most are API wrappers with a ChatGPT logo on them, sold at custom-solution prices. A guide to reading an AI pitch without being impressed by the demo.
Recent decryptions
When AI floods Linux kernel security: the signal nobody's really commenting on
In May 2026, Linus Torvalds declared the Linux kernel's private security list unmanageable, flooded with AI-generated vulnerability reports. A documented case of externalized automation costs, and a test for critical thinking about AI.
Browse by topic
Reality vs hype What AI actually does, the hype cycle, false promises. The site's thesis axis: separating what's sold from what's real. Mechanics How it works and how it learns: algorithms, machine learning, deep learning, LLMs, vision, reasoning. Not just LLMs. Economics Cost overruns, quotes, ROI, build/buy/API trade-offs. What the market leaves out of its pitches. Risks Hallucination, bias, quality, confidentiality, trade secrets, GDPR and the AI Act. What they don't show you in demos. Sovereignty On-premise, open source, hardware (GPU/APU), running AI yourself, US vendor dependency. Taking control without kidding yourself. Usage When AI helps, when it doesn't, how to decide. The 3 questions to ask before adding AI anywhere.
Reading path
- 1 Reality vs hype
Start here: the thesis axis, what AI is and isn't.
- 2 Mechanics
How it works, without equations.
- 3 Risks
The real risks, not the fantasies.
- 4 Economics
The real cost, the real trade-offs.
- 5 Sovereignty
Taking back control.
- 6 Usage
Deciding when it helps, and when it doesn't.
Why this site
I spent years watching executives make AI decisions with only vendor pitches and press articles as their compass. The result: projects launched on false premises, wasted budgets, and growing distrust toward anything labeled AI.
This site is the counterpoint. Not a list of best practices, not neutral pedagogy. We take apart what deserves to be taken apart, acknowledge what genuinely works, and cite everything.
— François Aichelbaum, Privateer