Ray Dalio: "Using super-powered AI, we might just destroy each other"

In February 2026, Ray Dalio launched a digital version of himself - a chatbot that should answer questions about markets and the economy in the logic of the founder of hedge fund Bridgewater Associates. One of the first testers of "digital Dalio" was Oninvest correspondent Anna Krasnova. Thanks to this, she recently received an invitation to a closed online Q&A session with the real Ray Dalio.

At the meeting, the investor talked about how AI is changing decision-making, why people need an intelligent opponent rather than an obedient assistant, what is happening to markets against the backdrop of the technological boom, and what risks he sees in world politics. Oninvest recounts his main statements at this meeting.

The international arena is a changing order. The scenario is always the same: the dominant powers set the rules of the game after a war, and through that determine the whole order. Over time - often a period of about 80 years - the situation begins to change, and this coincides with the end of the debt cycle and other processes.

I remember a time when the U.S. only had to hint to other countries what it wanted and they would immediately fulfill it. They were well aware that the consequences - whether they were rewarded or punished by the superpower - would be very tangible. Now we are witnessing a fundamental shift in the perception of this power, including in the situation around the Strait of Hormuz.

Comparisons with the British Empire are apt. The turning point of its withering was the Egyptian seizure of the Suez Canal, when the British simply could not regain control.

We should expect high uncertainty, massive change and, as a consequence, enormous volatility. The question is how to play it: the temptation to think in terms of "what's about to go off" in a turbulent environment encourages people to concentrate risk dangerously. Look, for example, at the market capitalization of AI companies - it has become absolutely dominant, investment portfolios have become highly concentrated in this sector.

The world currencies that have existed since 1750 have either been destroyed or devalued. About 85 or 90 percent of them. That is the very nature of the process; it is inevitable.

The main symptoms leading to currency depreciation or collapse are directly related to debt. A debt asset is essentially a promise to pay a certain amount of money. But no one guarantees how much that money will be worth in the future. As a result, the currency loses its purchasing power.

I think that the huge debt problems, the geopolitics that are changing the way countries think about what assets are safe and comfortable to hold reserves in - all of this is having a tremendous destructive effect on the value of dollar debt and the dollar itself. And now you can see the reflection of this process even in the dynamics of the markets.

The times when you could just scroll through the markets in your head and make decisions are in the past. I think most investors are just now starting to face that reality, because AI has only been on the horizon for a few days. So right now, I think people are using it in a rather chaotic and haphazard way.

But in the future, each person will have to develop their own AI systems. It will be necessary to build thinking in partnership with the right AI or a whole set of AI tools - to work in synergy, where the best qualities of human intelligence are revealed. I believe it's still necessary to have your own insights and beliefs, but it's important to have them questioned by a trusted partner - a trusted AI partner. With its help, you can build decision-making systems. Simply put, you need to be at the forefront of these technologies or you will lose. You're just going to fall behind.

If you interact - especially now, with large language models (LLMs) - do so not for ready-made answers, but for discussion and exchange. Seek to understand the cause-and-effect relationships that drive processes and form principles for working with them effectively. This is the real path.

It seems to me quite possible that the tremendous growth in intelligence of machines will be accompanied by a complete lack of worldly wisdom in humans - so much so that, using this superpowered intelligence, we may simply destroy each other.

This article was AI-translated and verified by a human editor