I listened to an episode of this week V for Valentine — and as is often the case, one sentence stuck with me. Not from a grand geopolitical argument, but from a seemingly casual sidestep about AI. This blog is about that sentence, about how to critically assess a source, and about the difference between “something works” and “something is well built”. With respect and admiration for the makers — because they deserve it.
An episode worth watching
The episode in question is #268 — “The downfall in a nutshell”, in which Robert and Boris visit the hot spots of the moment in more than fifty minutes. The Japanese Yen Carry Trade and what it does with your savings, the Far East as a mirror for the Netherlands, the question of whether the Amsterdam riot police is even ready for what is coming, the European passport and the digital euro (CBDC), and what Alice Weidel has to say about the EU. And somewhere in the last quarter: how AI tools are changing the future for the individual.
What I appreciate about V for Valentine is exactly this: they tackle topics that are rarely or never covered in the mainstream media, and they do so with context instead of sugarcoating. You don’t have to agree with every conclusion to recognize that the agenda is valuable. People are working, reading and asking questions — and that’s more than you can say about much “mainstream”.
But – nothing human is alien to us
And yet. This is where the stamp above this blog comes into play. Because there is a human reflex that plays tricks on all of us, whether we look at the mainstream media or an alternative channel: we accept something more quickly if it fits into our narrative — into the image we already have of the world. Does it match what I already found? Then it slides in without friction. Does it clash? Then we dig our heels in and look for the mistake.
That is not a fault of V for Valentine, nor of their viewers. It is a blame on our wiring. Confirmation bias has no political color and no media preference. That is precisely why the healthy attitude is the same for every source: appreciate what is good about it, and at the same time continue to assess it critically. Who is the source, what is the importance, what is not said, and – the most difficult – Maybe I want to believe this because it fits so well? A “fact check” stamp on a poster is a wink, but the underlying practice is deadly serious.
And I saw exactly that reflex in a small, sympathetic moment in the AI chapter. Not about politics — about technology.
“That’s Just How It Works Now” — the AI chapter (from 46:00)
Round 46:00 the chapter begins “AI tools are the future for the individual”. The gentlemen talk about privacy, about data that always passes through third parties, about how as a user you no longer have any control over your own messages. Recognizable and rightly so. And then tells Robert an anecdote that many makers will recognize (around 48:21):
“I wanted a booking plugin for a WordPress website this week, and everything that was there just didn’t look right. So I said to Claude, ‘I like this booking tool — is that WordPress?’ No, that’s WordPress with a custom-made React tool. I say, ‘I want you to make that tool into a WordPress plugin.’ I think that took me two hours.” cost, and now there is an ultra-sleek booking tool on the website. And that costs me nothing, while those plugins cost fifty dollars per month. Just built it myself, from my own GitHub. That’s just how it works now.”
(Excerpt from V for Valentine #268, ± 48:21–49:05, lightly edited from the automatic subtitles for readability.)
And it gets even more beautiful. Boris hooks up with a holiday story: a day off, bad weather, and instead of letting the kids play games he said, “Okay guys, we’re going to do an AI project.” Together they built a website with games and scorecards. Honestly? That’s exactly what we want to see. Children who build something with technology instead of just consuming it. Chapeau — that makes me happy.
The bulldozer
The trigger for this blog was Robert’s story. He approaches it with one consumer view, and that is completely understandable: what do I want, what does it cost to buy or rent — that’s a lot — and now I might as well let AI build it myself. As an argument, that is inspiring, and it is true: the threshold for making your own has dropped spectacularly.
And yet I want to put an image next to it, because this is where that human reflex returns – this time not with news, but with technology. Compare it with construction:
Give a skilled homebuilder — someone who knows exactly how to lay a good foundation and has until now dug out the ground with a shovel — a bulldozer, and he will do his job faster and better than ever. Give that same bulldozer to a random person, and you will probably end up with foundations in places where it would be better not to build a house: in the middle of a river, straight through sewer pipes, in places where an experienced builder would never dig.
AI is that bulldozer. A phenomenal tool — we use it ourselves every day and with great pleasure. But the tool does not determine the quality of the result; the hand around it and it foundation underneath do that. “It works” and “it’s in the right place” are two different things. And nowhere is this more evident than with something as seemingly simple as a booking tool.
A booking tool: the consumer route vs. the safe way
Such a self-built booking tool works — it looks sleek and costs nothing. That is precisely why we quickly accept it: it fits perfectly into the narrative of “free, fast, and completely my own”. But now ask the questions that a professional does ask, and that are easily skipped in the euphoria:
- Where are the booking details? Names, email addresses, telephone numbers, perhaps payment details — on which server, under whose control?
- Does the tool secretly talk outside? Many “free” components load third-party fonts, scripts or analytics — thus leaking visitor data without you noticing.
- Is it GDPR-proof? Processing agreement, retention periods, consent — arranged, or not there?
- Who can join? Authentication, permissions, and what happens in the event of a leak or crash without backup?
This is the same pitfall as with the news, only in code: something feels good because it fits our story, so we don’t check it. “It works” is not proof that “it is safe”. A tool that takes two hours to install may have created a data breach in those same two hours that you only discover months later.
This is how we build it at Brainfusion. We don’t start with the tools, but with the architecture. Self-hosted, without dependence on big tech, and designed in such a way that your data stays with you and does not end up unnoticed by a third party. We use the same AI bulldozer — but with the foundations in place: security, privacy, backups, GDPR, and ownership baked in from the first line. You get the speed of “building it yourself”, with the thoroughness of someone who knows where he is digging. A booking tool, a scorecard app for the children, a customer portal — the same cheerful building drive, only in such a way that you will still be happy with it three years from now.
Do you want something like this built professionally and safely? Please feel free to contact us — we are happy to think along with you, from a loose idea to a solid solution.
See for yourself — and support the makers
This blog started with an episode that got me thinking, and the least I can do is send you to it. Above all, judge for yourself, critically and with an open mind — exactly as it should be.
- 📺 Watch episode #268 on YouTube (and subscribe the canal).
- 🎧 Or listen via Spotify and Apple Podcasts.
- 💬 Join the conversation through them official Telegram channel.
- 🔥 Do you appreciate the work? Become a Bossy Boss and support the podcast via vforvalentine.com/join-the-tribe — then you immediately get access to “the Xtra”, the extra content for supporters.
Image: promo poster of V for Valentine #268, edited by us with a wink stamp. All credits for the episode and the visuals go to the makers of V for Valentine.
mensen zouden zich wat breder en dieper in bepaalde materie moeten inleven, eerst een plan maken en dat toetsen of het bestand is tegen avg en andere regeltjes. er worden dingen gebouwd en de wereld ingesmeten …
En de Chat opties via een webbrowser , met of zonder abonnement , maken super veel ” inschatting fouten” , lokale LLM’s zijn “stemming” gevoelig ( by design? ) en kunnen alleen door een persoon met kennis regelmatig bij gestuurd worden .
Luister dit korte clipje :
MyRobot is a Time-Saver – https://fountain.fm/clip/gcQmNoJvp6zX61rH7ZRm
je opbouwende kritiek heeft gerusulteerd in een uitzending waar AI volop de aandacht kreeg en blijkbaar willen ze de wereld vertellen dat je geen rekening hoeft te houden met de punten die je aangeeft. Heb je een antwoord ontvangen van Boris of Robert of alleen van een medewerker ” ik geef het door” ? ( mijn ervaring eerder ) . Blijkbaar vinden ze het verhaal hoe zei het vertellen belangrijker – of zoals regelmatig gebeurt ” laat ik mijn AI even laten kijken … ” dan weet je dat je onzin antwoorden terug kan krijgen , laat staan als je Vibe Coding gaat doen zonder concept of programmeer ervaring – “What can Possibly go wrong?”