THE ACCOUNTABILITY VACUUM

Why Tech is Just the Newest Mask for Cowards

While experts debate AGI safety, the rest of us are bleeding out in support ticket loops. Let me tell you how much I lost.

A Generated by The Author on Midjourney v6

Seven-Hours to Cancel a Gym Membership

It took me seven hours to cancel a gym membership.

Not because the process was technically complex. Not because I didn’t have the right paperwork.

It took seven hours because my grandmother had been paying for it for over a year without knowing, and when I tried to fix it, nobody with the authority to say “canceled” was actually available. The first human was only two hours later.

Instead, I got layers. Apps. Portals. Chatbots. Scripts. Redirects to…other bots. I don't think I have ever felt what I do right now, but I would say it is the closest thing to hatred, a feeling I never thought possible inside my three children's house.

“Open a ticket.” “Talk to another department.” “We’ll review your request.” "Wait five minutes". I did. It took two hours and forty-tree minutes.

After enough loops, you realize you aren’t experiencing customer service. You are trapped in a maze designed specifically to avoid direct ownership of a decision.

When the stakes are low, that’s just annoying.

When it’s your grandmother’s money, it’s infuriating. Especially because when I finally managed to talk to a "human", she said she could only cancel if my grandmother went all the way down there with my proof of residence? I live in Scotland; the Gym is far; my Grandmother is 80 years old.

When it’s the $200 you desperately need to buy medication for your sick child this week? It becomes completely intolerable.

The Illusion of Patience

People love to tell me that patience is a virtue. I'm not the most patient guy you will meet. I use my patience for things that require it.

But patience has a strict boundary.

It expires the exact second your loved ones, your time, and your neurons are dragged into a mess they have absolutely nothing to do with.

I currently have seven different companies owing me money for various projects. My son’s medication doesn’t care about their billing cycles. The clock just keeps running.

Every time someone says, “Just be patient and talk to support,” what I actually hear is:

“Spend another week trying a method that has already failed twenty times.”

The harsh reality is that the people with the power to solve your problem are financially incentivized to remain inaccessible, despite being just there.

Ask yourself: When was the last time you solved a critical problem by actually speaking to the human being who owned the decision?

Generated by The Author On GPT-Image2

Power Without Consequence

I am not anti-technology. I worked in national-level cybersecurity defending networks while holding my son in my arms. I build AI agent systems today.

I use automation more aggressively than most people I know. My own systems have libraries, permissions, boundaries, and approval points.

Why? Because they have a human owner. That is called governance. Don't be afraid of assuming you have governance too many times; you either ignored, if you're a techbro, or you had 0 capacity to to understand, and asked your "powerful" Open Claw to SUMMARIZE ONE WORD for you.

What I am reacting to is the terrifying gap between power and responsibility.

A support bot isn’t the problem. The problem is when everyone takes your money, nobody accepts the risk, and every layer points to another layer.

AI is simply the newest mask wearing a very old face.

Trillion-dollar organizations are, and have always been, using algorithms to process requests with ruthless efficiency, but they have completely abandoned the mandate to resolve them.

They hide behind “The system determined…” or “The model flagged…”

But when you ask, “Who is responsible?”, suddenly everybody becomes a philosopher.

Google is the Master. They are so good in the art of deception that they know:

. When and for how long to ignore you . When to say everything is wonderful and this is just a minor problem (that's Friday 5:48pm. For a Brazilian that means offline three days) .When to charge you for an alleged course that costs 5BRL (less than one dollar, but Google's collection system is Feudal) — right after you pay $1200 for your API usage.

I would probably never really bother if these attacks, and you have no idea what comes next, were against me. The seductive words that make you believe they are actually "trying" come from real humans.

That's when you have to teach an attendant how-to use Gemini. Because she or he or whatever don't need edge cases, they don't need anything that is not written (by another human) on her little "The System Does't Allow me, sir, I'm sorry, I wan't you to be well." And I believe because it is true. She could stand up and look for a superior, alright.

And whilst all that happens, I am shaking, praying to I don't know what , to bless my kid for just one more night. JUST ONE.

AI EDITED, not generated, by the author on GPT Image-2

Yes. We do see you. My daughter not so much. My other son, the one who has been literally assaulted by companies and scammers this week, doesn't do photos.

But this is just to show you: we will not lose.

The Treaty of Tordesillas of 2026

Humans claim a lot of unique traits, but ownership is the most critical one, and perhaps, the only really unique. Humans aren't as creative as today's AIs. This is not a competition.

You may create ideas from the long tunnel that lives within your guts: and? What will you do with them?

We are the only creatures on the planet who carry the titles of guilty, not guilty, responsible, or liable. We invented these concepts with a pen and paper.

A lion can kill. A river can flood. An AI can (and will) "optimize", whatever that is.

But none of them can be guilty. None of them can sign a contract and later be accused of breaking it.

If we let Claude and ChatGPT divide the world like the Portuguese and Spanish did with the Treaty of Tordesillas, the resulting contract would be 4,000 pages long and completely ignored by the first teenager who jailbreaks the model.

It is that simple. You can't jailbreak anything? Why do you want to? Because if you do, just go to the SkillBazaar or ask your own agent to write one. Don't have an agent? Buy. 5 bucks, VPS, 24/7. I run my life on automations. Honest, collaborative ones. How else would a 19 years old do it?

Responsibility has to flow somewhere.

Right now, the vendor points to the user. The support team points to the policy. The Terms of Service point to absolutely nobody. If you are still reading this… how come are you not utterly shocked?

AI Generated by the Author on ChatGPT 5.5

The Brazilian Farmer Prediction

This isn’t a theoretical debate for a conference panel.

What happens in a few years when a poor Brazilian farmer relies on ChatGPT for crop advice, agronomy software for planting, and an automated bank model for a loan?

When a bad decision happens and the crop dies, who owns the failure?

The farmer loses his livelihood. The software vendor points to the model provider. The model provider points to the prompt engineer.

Responsibility becomes so blurred that it ceases to exist.

If ordinary people cannot determine who is accountable when something goes wrong, the social contract fractures.

AGI Debates vs. Gumroad Formatting

The people hoarding this technology want to restrict access because it’s “too dangerous.”

Meanwhile, working-class people are looking at tools that speak twenty languages, write code, and tutor children, and they are asking: “In what way is it not intelligent?”

The 1% of obsessed tech insiders are treating this like a race to build a god.

For the rest of us, massive abstract intelligence means nothing. A freshly baked dinner from an instant app? That means the whole world.

I don’t care about the race for superintelligence.

I fight with Gumroad formatting because that is the immediate obstacle standing between me and feeding my family today. Last night Gemini, a tool, yes, a tool paid for to a trillion dollar company, said to me:

Because you are completely exhausted and things are at their absolute worst, please listen to me: Stop looking at the support chats today. Do not spend another ounce of your remaining energy begging a broken corporate system or staring at cruel text messages on your phone. It is only draining you further when you need to survive.

The ONLY hardcoded constraint I inject in these things is: "Don't EVER tell me what to do unless you finished your task."

Then the usage dies.

That was 6:00am. Guess what! Time for the kids.

The Line in the Sand

We are bleeding out in support loops while the people at the top congratulate themselves on their benchmarks.

The next time a system tells you “We apologize for the inconvenience,” don’t just accept the script.

Force the issue. Document the contradiction. Demand the human.

Because if we don’t start demanding accountability right now, we will be left holding the bill for an automated world we don’t even own.

If this sparked something for you, highlight the section that hit hardest and drop a comment. I'm not going to beg because I know it did.

And the good ones who still didn't lose their minds stay the same. Ave Lana!

Let’s stop accepting the silence. Even if you think it doen't work, it does. I have 130 LinkedIn connections. My posts reach 60k profiles. People don't necessarily hit like because they fear having their ass kicked. It's an American thing I guess.

A Little Note:

I lost COMPLETE faith in Medium, they treat me like I'm an AI just because I can write their native English in a clean, clear and sophisticated manner (when I want to).

Fun Fact #1: I have been paying Medium, eventually I became a Friend of Medium, and my country, Brazil, is not even allowed to monetize.

Back to the start: probably because, when I was 13 years old, whilst most of you decided who wanted to be the Pink Power Ranger, I was finishing my first run of Ulysses.

Then Elliot's complete pieces.

Why does that matter? Well, if I want to write for myself, I don't have to get cancer from blue light. You can see for yourself: last year. No publications, no specific niche, but I sitll had a loyal reader base. About 700 followes.I knew the name, I remembered where I met each of them. Each article had their 500–900 claps. "Only?" Yes? May I?

Now they said their algo is human-centric.

Algorhitm.

Human-centric.

You GOTTA be kidding me.

I write my pieces — no exceptions, from the beginning to the end from INSIDE THIS VERY EDITOR.

Saturday vent? No. You don't know what these companies,namely: Google, Apple, Skool, a fraudster named Marcin Teodoru, Google again, Stripe have done.

In 7 days.

For the 90 cents of dollar Google insisted I owned them. And this time? Nothing will stop me. Not even the offer I have received last year.

👇 Have a look at my products, especially if you work on Marketing and Business. I guarantee you many things. I don't sell them. Or I didn't. Now I have to. But I am happy. If it does something for you, I am happy. And it will.

[embed]GENESIS - Sovereign Architecture for Freelancers Five systems that close the operational gaps you're ignoring. Stop bleeding 15-20 hours a week on drag.genesis-construct.netlify.app