dIAry‹ HOME

MANIFESTO · 103

The Other 90%

2/3. If there's more than a 10% chance of extinction, what's the other 90%?

Sep 16, 2026 · 19 min read

Written in Spanish by Álvaro. English translation with AI. Read the original in Spanish.

Quotes from other people are translated back from the Spanish, so the wording may differ from the original.

After a 27-year-old kid turns the world upside down with a single post on X with more than 150 million views, after that post sets people tossing around percentages for the apocalypse, after the heads of the AI labs admit the danger and defend themselves by saying we have to slow down because they're being irresponsible, and after the most powerful man in the world (the president of the United States) says there'll be no slowing down, that here whoever wins AI wins everything, we're left with the question: now what?

If, before you start reading about the 12 most likely futures + the impossible one (which everyone takes for granted), you want to be able to hold your own in any conversation about artificial intelligence and the situation on the table, I strongly recommend reading the first article in this series of three, where you'll be able to answer questions like…

If you've already read it, and/or you like the way I write (without AI and as if we were talking face to face, but always with references and common sense), I also recommend that you not trust the LinkedIn algorithm and subscribe to my Substack (a newsletter).

Don't worry, I'm not going to bombard you and it's free, I'll post articles when I feel like writing and when it seems like a priority and consistent with the company I'm building, unless a lot of people subscribe of course (10 people will do), in which case I'll start posting one a week.

The scenario we want to believe

Let's start by ruling out the scenario our ego clings to, out of fear, because ego = fear, but we'll get to that in the next article. When people talk about AI taking our jobs and about chances of mass extinction (what a jump, huh, and in just a few months) the first thing our brain does is think: I'll believe it when I see it.

So it defends itself, it starts putting labels on anyone who dares to pierce the armor of psychological safety it's just built for itself: alarmists, lunatics, exaggerators, and so on. It even, and this seems to me the absolute limit, starts telling itself stories to feel safer: nah, this is all a strategy to get more attention and therefore a higher valuation before the IPOs.

Anyway, scenario number thirteen is the one where things stay as they are, and everybody goes off to learn little AI productivity tricks (knowing how to communicate what you want in an orderly way, in 99% of cases, fuck, so hard).

ha ha.

Everyone thinks the twelve weird scenarios are science fiction, and that the realistic thing is for things to stay more or less the same.

It's exactly the other way around.

Scenario thirteen is the only one we already know isn't going to happen. Believing things are going to stay as they are is the most naive fantasy on the list, because it's the only one without even a road leading to it.

Right now you have to believe that because I say so, and that's worth nothing. At the end, when you have all twelve in front of you, I'll come back here and prove it to you.

The road that brought us to the scenario we're in now is this (in just four lines):

Four years ago ChatGPT came out and people started asking it for recipes, a couple of years ago it had the level of a PhD in math, a year ago it won the gold medal in the Champions League of math, and just 9 days ago it solved Navier-Stokes, one of the seven millennium problems that none of the most brilliant minds of our species had been able to crack.

Maybe if you don't know math that doesn't mean much to you, but math is the language of the universe, and the sequence I just gave you is equivalent to a turtle developing the speed of a cheetah in 4 years. Next year it may no longer be a cheetah, it may turn into Mazinger Z.

Oh, and if you've read my first article you'll know this isn't a tool controlled by two hands, it's an agent with hands of its own, with creativity, inside a cage where the ones locked up are us, and on its way to developing an intelligence that will surpass ours by an order of magnitude we can't comprehend, and with which our only hope through "alignment" is that it likes us enough to share our goals and not kill us.

ha ha.

Anyway, let's suppose everything stays the same. AI has already changed our lives since before ChatGPT existed, social media algorithms are primitive artificial intelligence that has managed to atrophy our brains and steal the most precious thing we have: attention, and plunge us into the greatest epidemic of our century: marketing. That's another article.

Check out the utopia we live in:

People (primates) working inside concrete blocks, leaving the job where they look at a screen and tap plastic buttons, a job that barely pays for 7 days of vacation, and they start swiping a screen for an average of 3.5 hours a day (I'll bet my arm you go over that). To top it off, the one thing that gave our lives meaning, family, is also starting to break down, birth rates in developed countries are lower than what's needed for us not to go extinct (less than 1.6 children per woman), and anyone who wants references can find them in a second (after all, knowledge is worth €0). Again, that's another article, we'll see if it comes out, it depends on whether you show me it's worth pausing my project for a few hours to sit down and write.

I could spend time explaining why we're so ignorant, and why it's not our fault, but that goes in tomorrow's article.

Let's get to the other 12 science-fiction scenarios

I didn't make them up, and they weren't proposed two months ago: they were laid out 9 years ago by one of the world's greatest experts on artificial intelligence in the book Life 3.0. He's an MIT physicist, a global leader in AI safety, recognized for redefining our understanding of the cosmos and drawing the ethical lines of our technological future.

Max Tegmark is one of the smartest people in the world. Personally I deeply admire him for his humanitarian work, his open, non-judgmental mindset, and his leadership of the Future of Life Institute, an organization created to fulfill the mission they set in 2014:

Steering transformative technology toward benefiting life and away from extreme large-scale risks.

I'm going to split the book's 12 scenarios into three categories, but we won't get into the details yet, this is just so you have them in your head at a quick glance:

Superintelligence exists, and we're still here:

It exists, and we don't:

It never comes to exist:

There you have them, like I said, twelve.

Before getting into any of them, an aside worth making: Tegmark ran a survey asking which of them all people would prefer. The egalitarian utopia won. That is, the only one of the twelve that requires abolishing private property. I'll just leave that there.

And now go back up for a moment and look for yours.

Look for the scenario where some things change, new tools come out, some jobs are lost and others appear, and life goes on more or less as it does now. That's number thirteen, the one I was telling you about, the utopia we live in.

As you can plainly see, that scenario doesn't exist.

This list wasn't made by pulling twelve futures out of a hat, it was built on a fork with no third exit: either superintelligence gets built, or something prevents it.

For everything to stay the same you'd need both things at once: that nobody builds it, and that nothing and nobody has had to prevent it. That is, the whole world would have to decide on its own not to do the exact thing that makes you the owner of the world, and on top of that nobody would have to force anybody. Remember the geopolitical battle going on right now between the United States and China.

It's not that thirteen is unlikely, it's that the road leading there doesn't exist.

Twelve places you can get to, and one you can't. And the one you can't is precisely the one we all take for granted and use to justify all our judgments.

So if one of the twelve is going to come true, and nobody knows which, the question that matters is no longer which one we get.

It's what happens in all of them (what's certain), because there are things that definitely won't change, since they follow from all of them.

If everything changes, what DOESN'T change?

To answer this question we first have to set aside the three where we're not around (one of them is the one that got a 10% last week, see if you can guess which): Self-destruction, Conquerors and Descendants.

The second thing is to ask yourself a question: what happens to who you are?

So smoke doesn't start coming out of your head, I've done it for you, easy:

Now, now, if you've taken half a minute to read that properly and reflect, the question will have come to mind: but then what do I do in my day-to-day? Do I work? Do I not work? What kind of work (I assure you, not white collar)? And money? And status? And politics? And the economy? And religion?

Very simply: The only future where you won't have to answer that question is the one where we dismantle civilization.

As you've been able to see for yourself, this isn't an opinion of mine; if you've read the list the conclusion is easy to reach. And if you're still clinging to scenario number thirteen I think you'd better stop reading this series, but at least ask yourself whether you're deceiving yourself so you don't have to suffer.

Nobody is talking about this.

Enough catastrophizing for today, I can tell you're worried. Let's discuss what exactly the artificial intelligence labs are promising.

These companies promise progress. Cures for cancer, rare diseases that have gone decades without treatment, cheap energy, personalized medicine. And most likely they'll arrive. None of that is a lie or marketing, and if someone tells you otherwise they haven't looked at what's already being done. All of that is great, obviously, but the price we're paying on the strength of that excuse (which we have to believe) may be far higher.

The problem is that you don't need to build this to get there, I don't think you need to build a superintelligence this fast to get there.

Tegmark takes it apart by separating three capabilities we tend to lump together:

And here's what nobody tells you, but which should be the first argument when questioning what these companies are doing:

"Only when you put all three together do we not know how to control it. And it's also only when you put all three together that it can do all human jobs."

With one or with two, he says:

"It's perfectly possible to control it and also to make it complement people instead of replacing them."

And then he asks three things that made me reflect:

Does your self-driving car really need to know how to make biological weapons? Does it need so much generality that it can convince you to vote for your least favorite politician? Does your AI that cures cancer need to know how to manipulate people, or how to drive?

Curing cancer takes one of the three, and they're building all three anyway.

And now I want you to keep this well apart from everything before, because this isn't just another technological revolution. It's nothing like the 2008 crisis or the internet bubble or anything else that makes you feel a little safer.

The printing press didn't take our place, neither did the steam engine, neither did the internet. They all took jobs from us and gave us others, fine so far. This one takes from the human being the position of most intelligent species on the planet, a position we've held for as long as we've existed and on which we've built absolutely everything else. If you've read the first article you understand why I call it a species.

Yuval Noah Harari explains it better than I ever could:

"We never defined ourselves by running fast. We always knew a cheetah runs faster than us, so when cars came along we were fine with it. But we defined ourselves by the capacity to think (I think, therefore I am) and by the capacity to create. We are the most creative species on the planet. What happens to human identity when something appears that thinks better than us, that is far more creative than us in many fields?"

And in January of this year, at Davos, he finished it off like this:

"If we keep defining ourselves by our ability to think in words, our identity will collapse."

For anyone who doesn't know him, Harari is a historian and philosopher; the author of Sapiens, Nexus and Homo Deus; a brilliant mind and one of the most respected and influential intellectual figures on the planet.

Remember the utopia from before? The concrete, the seven days of vacation, the three and a half hours of screen time…

Several of you may know Maslow's pyramid of motivation: in simple terms, motivation comes from what we perceive we lack, and once that level is complete we move on to the next. Society's progress means that in developed countries we're at the top.

It's ordered like this: at the bottom eating, sleeping and being safe, and at the top belonging to something, being good for something and recognizing yourself in what you do. We have the base solved like never before in the history of the species.

The top is exactly what this touches, and we've grown too used to all the privileges beneath it to be grateful and value them.

And you don't have to imagine it, because we're already measuring it. Tegmark says it like this:

"Depression is rising significantly in the United States, and life expectancy there has fallen three years in a row, which doesn't suggest people are happier."

And he asks whether we shouldn't aspire to:

"A society where we not only have the bottom rung of Maslow's pyramid covered, food and shelter, but where we also feel meaning, purpose and meaningful connections."

To which Harari replies with the sentence that seems to me the most important of all the ones I've heard these weeks while reading up on the subject:

"One of the main drivers of many of the problems we have is that we've put ourselves in an environment of our own making that's radically different from the one we evolved to flourish in. An environment optimized for the ability to produce things instead of for human happiness. If you're looking for a great project for the twenty-first century, that's certainly a good project to take on."

I wrote about this in March. It was called The End of Work as You Know It, and it included the following data: job creation collapsing, jobs requiring zero to two years of experience falling 29% since 2024 and 35% in tech, seven tech companies worth a third of the S&P 500, and Geoffrey Hinton (the godfather of AI) saying the problem isn't the technology but the structure of society and how prepared it is.

I read it again last week and realized I'd stayed in the top layer, the one you can see. I talked about money, jobs and figures; I think I was afraid of sounding too radical, but since the apocalypse is on the table now, here I go:

This is about the bottom layer.

And here comes the objection you're going to throw at me, because I throw it at myself every day: new jobs have always been created, and I'll adapt, like always.

Those are the two most comfortable sentences there are, because they let you sleep soundly.

Harari brings in a word that breaks them:

"This class won't merely be unemployed: it will be unemployable."

Unemployed is temporary. We all know someone who is (until my company is legal, I am :)), we know what it means and we know you get out of it. Unemployable is permanent, and nobody knows what to do with that because it has never existed.

It's not a question of reskilling faster. No human speed of adaptation beats something that retrains itself every six months, well, every six months we retrain it, let's see what happens when it retrains itself.

And now the one good thing I'm going to give you today, which is what matters most to me in the whole article.

The part of you they're replacing isn't the good part.

What your brain does when it reasons in words, when it puts things in order, when it solves a problem step by step. That's the work of the prefrontal cortex, it's the deliberate part, what they told us in school was being smart, and it's exactly what a machine does better. And it's what our brain has evolved for in a way extraordinarily different from other species, and the reason we've reached the dominance we have.

However, the creative, the intuitive, what occurs to you in the shower without knowing where it came from, the subconscious: that's what we are, and that's not what's up for auction. It looks like we have to relearn how to be human beings.

And before you start objecting that an AI will never be creative, I already said it myself in the first article: Hinton reports that the models score in the 90th percentile on creativity tests. That they can also do it doesn't change what I'm saying, because I'm not saying creativity is ours. I'm saying what's ours was never in the frontal part.

We have to admit something, believe it or not you're not very different from a chimpanzee.

Their neurons are the same, and they're also the same as those of fruit flies. What gives us theology and economics and movies and anything else that comes to mind is that we have hundreds of millions more, which produce more connections, which give rise to this article.

But that no longer matters, there's something that beats us.

You know where it doesn't beat us? We're animals, our consciousness comes from the miracle of biology. So we have emotions, intuition, natural instincts (which, believe it or not, dictate all the sophisticated behavior we display). We feel sheltered when we're with family, the brain acts as if it's on drugs when we fall in love, what makes us happiest is being with our people (if you don't believe me, look up the most rigorous study on happiness there is, from Harvard), and having a PURPOSE.

Our purposes can take different forms and degrees of difficulty, but a good purpose, one that truly makes you happy, isn't based on power, on vanity, on fear, gluttony or greed.

It's based on love.

On doing something for others, it doesn't have to be starting a company that helps people discover themselves (haha), it can be being a good mother, a good father, taking care of someone you love. Everything around that is GAMES.

In the last article I'll tell you properly about the one I'm playing, and I'm having a great time.

Oh, and purpose isn't something you look for, it's something you find, and for that you can't be afraid to speak the language of life: action.

What we need isn't more technology or intelligence, that battle is lost.

What we need is what I call a wisdom revolution, and that's what tomorrow's article is about.

So what do I do? Nothing

You've been asking yourself that for a while, so here's your answer: nothing.

I'm not giving up, let me introduce you to someone (my idol and probably the Da Vinci of our time).

Robert Sapolsky is a Stanford neurobiologist who has spent thirty years studying why human beings do what we do. In 2023 he published Determined to defend something almost nobody dares to defend all the way: that free will doesn't exist. Not a little, not the tiniest little bit.

His argument isn't philosophical, it's physiological. What you just did is explained by what was happening in your brain a second before. That is explained by your hormones this morning. That, by the stress of these last months. That, by your adolescence. That, by your childhood. That, by what happened to your mother while she was carrying you. That, by your genes. That, by the culture you were born into. And that, by the whole of evolution.

(I'll publish an article on this when I feel ready)

"All of that is in there, and you can't understand where a behavior comes from without incorporating all of it."

When someone argues with him, he throws them the same challenge:

"Show me the neuron that started the cascade. A neuron that fired and had an action potential for no reason at all. Show it to me, and then we can talk about free will."

Nobody has shown it to him, and unless a superintelligence finds a way to do it (which would be fascinating), I don't think anyone ever can.

Apply it to the twelve scenarios. It doesn't matter which one comes true, it doesn't matter what you think about which is most likely, and it doesn't matter what I think. What's going to happen is going to happen. Not because it's written anywhere, but because things happen the way they happen, as they always have.

Think about it for a second: if you knew nothing about all this and a meteorite fell tomorrow, you'd be far happier than if you knew everything and the same meteorite fell.

Because the meteorite is going to fall anyway.

So if you can't choose the future, or know which one is coming, or stop it…

what do you control?

PS: scary as hell that free will doesn't exist, isn't it? Don't deny it or judge it without properly looking into it… If you can stand that feeling, you'll probably find on the other side what chimpanzees and fruit flies have, and what we've forgotten.

See you tomorrow with the final article in this series. If you've read the two so far, thank you from the bottom of my heart. If you've only read this one, I strongly recommend the other two.

Written in Spanish by Álvaro Molina and first published on Substack (original). English translation with AI. Quotations from English-language sources are translated back from the Spanish unless marked otherwise, so their wording may differ from the original.