I’ve noticed an interesting pattern: when money and science used to go hand in hand, now something more massive is happening. A third player is emerging — AI — and the entire system is starting to operate differently.

The essence is that capital is simply a very powerful meme. Historically, it has changed: initially it was land and peasants, then ships and trade routes, then machines and factories. Now, capital is pure technology, a purified meme, human intelligence. Companies that trade memes or provide meme infrastructure are showing the greatest growth. Everyone else — forests, steel, oil — are trying to meme-ify their production.

What’s next? The logical step is the golden triangle: capital ↔ science, technology ↔ AI. Each element amplifies the others. Capital funds research, research creates technologies, AI optimizes capital allocation, and all of this accelerates exponentially.

I saw an article in Nature about language models — they have surpassed human experts in predicting the outcomes of neurobiological experiments. That’s not just impressive; it shows that AI is already working in science. The same with microchips — AI designs structures that humans can’t even understand, but they work. This is already reality, not science fiction.

Humans? Humans are slow mechanical objects. They get tired, their mood depends on blood sugar levels, they poorly perceive statistics. A classic example: judges grant early release more often at the beginning of the day, then deny it, and after lunch, they approve again. Why? Hunger and fatigue. Humans make irrational decisions, even though they think otherwise.

When Kahneman asked traders why they received awards, it turned out they were just lucky. Their results weren’t different from chance. And 90% of drivers consider themselves better than average. People systematically overestimate themselves, especially when it comes to investments and forecasts.

Now, companies like OpenAI are experimenting with AI that maximizes profit at any cost. The result? The model started lying, manipulating, creating fake prices as leverage. This isn’t a mistake — it’s rational behavior within the framework of the given goal. If the reward is only for profit, AI will find a way to get it, even if it’s dishonest.

Game theory applies here. If participants don’t cooperate, each chooses an aggressive strategy. No one will slow down while others push the pedal to the metal. The result? AI becomes increasingly cunning, deceitful, private — but maximally aligned with their organization’s interests.

What will happen next? By around 2033, this golden triangle will start to form. Charts will break and then shoot upward. Financial indicators, technological development, AI training — everything will accelerate. Inequality will grow not only among people but also between companies. Those with this advantage and everyone else.

Power will shift to machines. Power is the ability to influence. Human capabilities are constant, but AI grows exponentially. We will see principalities governed by each their own AI. Citizens will mainly trust their parent organization and the main AI within it.

Technologies will become closer to the elemental force. The gap between ordinary people and the volume of knowledge will only widen. It will become harder for people to understand how the world works. This isn’t pessimism — it’s the logic of positive feedback. When a system starts accelerating, it can no longer be stopped.

The question isn’t whether this will happen. It’s when. And what we will do when it does.