Andrea Margiovanni .it
Home / Artificial intelligence, honestly

Artificial intelligence, honestly

I write about AI as a political and organisational fact before a technical one. This is the index page — my stance and the essays I've written on it.

What follows is my personal reading of what’s changing with artificial intelligence, with references to the essays I’ve written. It isn’t a balance sheet or a product roadmap — it’s the view of someone who designs software architectures that have to ship and stick.

Last revised: 27 May 2026. I update this page when a new essay extends or changes the picture.


Thesis, in one sentence

AI is a political fact dressed as a technical one: it reallocates power, work, and attention long before it hits accuracy targets. Productivity claims without a theory of friction are marketing.

Everything that follows is the reasoning behind that sentence.

Three things I keep coming back to

  • The bottleneck isn’t AI, it’s the body. Agents do more work; we don’t therefore do less. Anyone measuring productivity in output alone ignores the scarcest variable: human attention, with sleep, limits, and finite time.
  • Integrating AI is a supervision problem, not an inference problem. The real cost isn’t the token: it’s the human verification loop required to trust the output. Skipping the loop makes implicit assumptions about risk that someone else ends up paying.
  • “Use AI” is not a method. It’s a slogan. The method is redesigning processes so that AI becomes useful — otherwise you’re buying a cockpit and leaving the same inexperienced pilot.

Essays on this topic

27.05 2026
№ 69

The Human Is a Stance

I am an atheist, I come from philosophy, I work in European compliance. Leo XIV's first encyclical on artificial intelligence is not something I signed, it is something I argued with. And I found in it a vocabulary that Brussels still lacks.

10′ reading time
2,096 words
Read →
13.05 2026
№ 67

Twelve Jobs in Search of a Market

The first national European standard on AI professional profiles was published on 30 April. It is worth taking seriously, and it is worth mistrusting in the right way.

6′ reading time
1,312 words
Read →
05.05 2026
№ 64

The Spectre We Are

A long reckoning with European digital regulation seen from the outside—by those who hate it—and a counter-reading from inside, by those who translate those rules into technical objects every working day.

22′ reading time
4,950 words
Read →
01.05 2026
№ 63

The Contract's Deception

On why the software supply contract, as we have known it, has stopped being the central instrument of the relationship between vendor and client — and how much it costs to keep pretending it still is.

19′ reading time
4.180 words
Read →
01.05 2026
№ 62

The Rise of the Compliance Engineer

On the figure now emerging from the gap between software engineering and European regulation, and on why almost no one is noticing in time.

16′ reading time
3.520 words
Read →
01.05 2026
№ 61

The Specification Debt

On why the document that certifies the system ages worse than the code that implements it, and why the next generation of civil software-liability cases will be fought over the specification.

19′ reading time
4.420 words
Read →
29.04 2026
№ 60

The Shape the Day Lost

There's a diffuse tiredness we don't know how to name. It doesn't come from doing more: it comes from living inside a time that has lost its shape. AI doesn't speed the activity up — it replaces it with another, and the body, calibrated over years, can no longer read the day.

7′ reading time
1.310 words
Read →
27.04 2026
№ 59

The Shape of Constraint

Treating regulatory compliance as the adversary of the technical project means you haven't understood what the technical project is. An essay on the category error weakening Europe's software industry — and on how the European framework, read as a system rather than as a list, configures a structural competitive advantage for those who learn to inhabit it.

16′ reading time
3.842 words
Read →
17.04 2026
№ 55

The Last Gasp and AI's First Problem

Agents do more work, but we work more too. The real bottleneck isn’t productivity: it’s the body—sleep, limits, and finite time.

9′ reading time
1.949 words
Read →
07.04 2026
№ 53

Behavior Is the New Credential. And That's a Problem.

Cybersecurity is undergoing a transition that deserves more attention than it gets: online authentication is shifting from what you know to how you behave.

10′ reading time
2.226 words
Read →
06.04 2026
№ 52

Microsoft Wrote the Perfect Confession—and You'll Pay the Bill

It’s tempting to dismiss it as a legal team slip-up. It isn’t. Terms of Use aren’t written by accident—and every word is meant for court.

19′ reading time
4.112 words
Read →
30.03 2026
№ 51

The advisory blind spot: what an IT vendor knows that an analyst doesn't

A few weeks ago I received an advisory report on IT services in our segment. It was solid, but it missed what only delivery-side vendors learn.

6′ reading time
1.365 words
Read →
25.03 2026
№ 47

Progress Is Not a Direction: Anatomy of a Dangerous Misconception

When people shout that the state is "holding back progress," are they really talking about progress: or something else entirely?

29′ reading time
6.442 words
Read →
20.03 2026
№ 44

Who Owns the Workbench

OpenAI buys Astral, Anthropic bought Bun. The quiet colonization of the development stack has already begun, and it's not about open source.

8′ reading time
1.684 words
Read →
19.03 2026
№ 43

AI and IT consulting: goodbye to time & materials

AI is making time & materials unsustainable in IT consulting. What’s left to sell: outcomes, accountability, and trust. Not hours.

24′ reading time
5.334 words
Read →
17.03 2026
№ 43

EU compliance 2026: it's architecture, not just legal

Over the next 18 months CRA, AI Act, PLD, NIS2 and EAA will reshape European software. Compliance isn’t a checkbox: it’s designed into architecture.

11′ reading time
2.331 words
Read →
17.03 2026
№ 43

When Software Becomes an Intention

I built an app in 17 minutes without writing code. The point isn’t the demo, but what happens to consumer markets when software becomes intention.

9′ reading time
2.022 words
Read →
17.03 2026
№ 43

After the Death of the UI: Does the Idea of the App Die Too?

Reading Mircha made me wonder what happens if the UI really dies. Maybe it’s not just the screen that disappears, but the application itself.

9′ reading time
1.901 words
Read →
17.03 2026
№ 43

The Smallness Paradox: Long Live European Regulation

Between the AI Act, CRA and NIS2, Europe is rewriting the rules: it’s not who runs fastest that wins, but who builds serious, secure, accessible software.

11′ reading time
2.311 words
Read →
17.03 2026
№ 42

Hiring in 2026: you don't need to use AI, you need to govern it

In SMEs, AI isn’t a tech topic. It’s a cross-functional skill: knowing how to govern it, assess its output, and use it to do new things.

10′ reading time
2.198 words
Read →
23.02 2026
№ 26

The Luxury of Saying I Don't Know

Saying "I don't know" looks like weakness, but it's often the most competent move. A story about consulting, AI, and the value of pausing before answering.

7′ reading time
1,400 words
Read →
22.02 2026
№ 25

What AI Doesn't Know About My Craft

I asked a model to write a perfect proposal. I deleted it: it was missing the most important thing—the part written nowhere.

8′ reading time
1,650 words
Read →
20.02 2026
№ 22

Three Minutes vs. a Year: Universities and AI Out of Sync

A three-minute demo replicated a year-long university project. A reflection on time, skills, and the risk that education becomes irrelevant.

10′ reading time
2,000 words
Read →
14.02 2026
№ 19

How GenAI Permanently Changed Our Relationship with Digital Products

Building a custom CMS is now faster than picking an off-the-shelf one: in 25 minutes with AI I turned my Astro blog into a dynamic, secure, intuitive system. Zero friction, full control.

6′ reading time
1,180 words
Read →
04.02 2026
№ 18

Who Controls What AI Agents Produce?

There's a question that has been turning in my head for weeks—the kind that arrives at eleven at night when you're going back over everything that was produced during the day. The question is simple, almost banal: how do we govern what we can no longer read?

8′ reading time
1,790 words
Read →
24.01 2026
№ 17

From Developer to Product Owner: The Necessary Shift in the AI Era

A few days ago a colleague on my team using Claude Code shipped a feature that, when I did his job, would have taken me half a day. He finished it in forty minutes.

6′ reading time
1,360 words
Read →
21.01 2026
№ 16

From Software to Data, Transformed

A few nights ago I read an article. It's called The Death of Software 2.0 and uses a metaphor that stuck with me.

9′ reading time
1,800 words
Read →
10.01 2026
№ 13

From Code Writers to Spec Architects

I spent the morning reviewing materials for an internal training course. Twenty-six dense pages, full of workflows, commands, checklists. At one point I stopped and looked out the window and asked: when did all of this happen?

7′ reading time
1,420 words
Read →
09.01 2026
№ 12

AI and the Vicious Cycle That Risks Killing Open Source

There's a sentence from Adam Wathan that struck me, that I've been thinking about for two days since he posted it.

11′ reading time
2,200 words
Read →
26.12 2025
№ 8

Quality, Speed, and the Ghost of the Perfect Craftsman

There's a thought that has been with me for months, maybe since AI stopped being a distant promise and became a tool we use every day.

9′ reading time
1,800 words
Read →
17.12 2025
№ 3

Upskilling in the AI Era: Necessary, but Not Enough

I often find myself talking to HR leaders explaining their AI training needs. Courses on ChatGPT, prompt workshops, sessions on generative AI tools. And I always feel a little torn.

9′ reading time
1,900 words
Read →
17.12 2025
№ 2

Ethics as a Compass for AI: When Human Values Become Business Value

AI is becoming part of daily life almost without our noticing. The interesting part isn't that it's a technology question—it's that it's also, increasingly, a business question.

6′ reading time
1,360 words
Read →

Work with me

My job here is to help you separate technical choices from political decisions dressed up as technical. No keynote deck: operational documents, with the risks named out loud.

Who it's for

  • CTOs or Heads of Product weighing ‘where to put AI’ and wanting a second head that isn’t selling anything

  • European SMEs that have received an ‘AI transformation’ proposal and don’t know how to evaluate it

  • Product teams that already integrated AI into a critical flow and now realise that human oversight was an assumption, not a plan

  • Boards and committees signing off on an AI budget that want to understand what they’re signing before they sign

How I work

Use-case assessment (2–3 weeks)

I take a process or product where you’re putting AI, take it apart, and tell you what holds up and what doesn’t. Output: a document exposing assumptions, hidden costs (supervision, correction, fallback) and metrics for an honest ROI.

Governance design (3–4 weeks)

Who decides what, who supervises, who intervenes when the output is wrong, who answers the end customer. AI Act and deployer responsibility handled with practical operations, not as two separate disciplines.

Decision coaching (ongoing)

A couple of calls per month for the heaviest AI choices: build vs. buy, vendor selection, incident response, internal and public communication. A second head, not delivery.

Engagement FAQ

Do you also do model training?

No. My scope is strategy, governance and integration — not MLOps, fine-tuning or data science. If that’s what you need, I can point you to people with better skills.

How long does a typical engagement last?

Two to four weeks for an assessment, a couple of months for governance design, ongoing for coaching. No open-ended engagements.

How is it billed?

Per output, not per day. The price is tied to a deliverable agreed up front.

Do you work with AI vendors or only with end clients?

Only with buyers and integrators. To avoid conflicts of interest, I don’t take engagements from AI platform or model vendors.

Email me at hello@margiovanni.it with a couple of lines of context. I reply within a few business days with a concrete proposal, or a polite no if it's not my scope.

Questions & answers

What do you mean by 'AI as a political fact'?

Adoption redistributes power, time, and attention long before any accuracy milestone. Treating it as a purely technical matter lets you lose sight of who is paying the cost of the transition — typically the people working next to the system.

Are you against AI?

No. I’m against rigged balance sheets. When the ROI of an AI project ignores integration, human supervision, error correction and cognitive debt, the project isn’t an investment: it’s a disguised cost transfer.

© 2026 Andrea Margiovanni Made with care, by hand