Principled Thinking and AI Need to Go Together
What is the best approach to being effectively intelligent now that human intelligence and artificial intelligence are merging? Because I have been building computerized investment decision-making systems for decades and I am now evolving them to be as advanced as possible using the most cutting-edge technologies, I reflect a lot on that question and I am often asked for my thoughts. In this note, I lay out my thinking.
I have found that to be good enough, especially in investing, there is no getting around the fact that one needs to make decisions based on logical, understandable criteria that use both great human intelligences and artificial intelligences. I know through my experiences that even the most advanced artificial intelligences don’t have adequate enough insights to allow one to blindly follow them and that unique human understanding and insights are still invaluable, and that that is especially true in investing where value-added is a zero-sum game (so that, when it comes to adding value, what is widely known is of little value.) I have also learned how principled thinking is an essential part of the process. So, my purpose here is to explain my approach to principled thinking in artificial intelligent decision making, why I believe it’s essential for effective decision making, and how it can be super-powerfully enabled with contemporary and rapidly improving versions of AI.
I am not being theoretical about developing principled thinking and converting it into a winning computerized decision-making systems, as doing that has been—and still is—the reason for whatever success Bridgewater, my Dalio Family Office (DFO), and I have had. I believe that the path to success is best achieved by putting the best human intelligence together with the best artificial intelligence and that the way of thinking I am describing here is essential to understand and use in the new human/artificial intelligence era.
In other words, I have found that to be effective one has to think in a principled way and align one’s principles to one’s decisions ideally using both human and artificial intelligence. Principled thinking is the examination and systemization of one’s decision-making criteria rather than just thinking to make decisions. It is best derived by thinking deeply about the circumstances that one is encountering and the criteria one is using to make decisions given those circumstances and then writing these criteria/principles down so that, when analogous circumstances come along, you know what to do. Principled thinking consists of making descriptions of how reality works—the cause:effect relationships—followed by the criteria/principles for what to do in that situation. To the extent possible, these criteria/principles are back tested to see how they would have worked and then computerized and automated so that they work in conjunction with one’s regular thinking to see how they align. Doing this with one or more great AIs (as partners) is invaluable in both teaching each other and making decisions.
To be clear, these criteria are not best derived by looking at what would have worked in the past and assuming that it will work in the future—i.e., data mining—or simply asking an AI what to do. They are based on logical understandings converted into decision-making systems. They work like a computer chess game that makes moves independent of you, with the understandings behind the criteria it is using clearly conveyed and with you operating in partnership with it, so you and your AI partner learn and improve together. Your AI partner makes moves via these systematic criteria while you make moves guided by the criteria in your head, so that you can compare the moves and the logic behind them and then align them. For those criteria that can’t be systemized, you can use subjective qualitative ratings and discretionary overrides. The system always speaks to you, explaining its logic (which you can debate with it) so you can understand each other and align your thinking and the reasons behind your moves.
I make sure that my principles are timeless and universal, meaning that I test them as far back as I can and in every type of environment and every country to see how timeless and universal they are. In cases where they don’t work out, I study why to build my understandings of the cause-effect relationships and refine my criteria. Said differently, I build up a number of criteria/principles that are essentially statements of “if this happens, do this because XYZ” and I explain the reasons which are then conveyed in the output so that the reasoning is always apparent. That way, the systems’ outputs are both logical and understandable and the systems can process much more complicated relationships more quickly and unemotionally than one can in one’s head.
That is the process I developed and used in my 50 years of building Bridgewater and that I am now using at the DFO to take full advantage of newly available AI technologies, and it I want to pass along to you. What this process can now do in creating understanding of the timeless and universal cause:effect relationship and enhancing and systemizing whatever one is thinking is mind-blowing. I believe that you will either stay at the cutting edge of doing this or you will be uncompetitive. I look forward to continuing to explain and help you stay at the cutting edge to be hyper-competitive to the best of my and my team’s abilities.
The chess comparison has an instructive sequel. For about a decade after Deep Blue, a decent human paired with an engine beat the strongest engines playing alone. Kasparov called it centaur chess and thought it was the future. Then the engines kept improving until the human half of the centaur started subtracting value, and today overriding the machine almost always makes the move worse. In chess, the partnership window eventually closed.
Investing is the rare game where that window stays open, and the reason is in this essay. Chess never changes its rules, so a machine trained on the game's past faces the same game forever. Markets change their rules precisely because everyone's machine learns them. The moment a pattern is widely held it stops paying, which is the zero-sum point made here. So the human contribution shifts. Out-calculating the engine stopped being possible years ago. What remains is recognising when the regime has changed and the past has quietly become the wrong training set. Causal principles survive those moments. Patterns mined from history die at exactly the moment everyone owns them.
Writing the criteria down in advance does one more thing that deserves mention. It makes you auditable to yourself. A principle on paper before the event can be tested and scored. Reasoning assembled afterwards is just a story you tell about yourself.
I would like to differentiate the understanding of risk and uncertainty in this context: risk is when circumstances repeat and you can learn the odds; uncertainty is when there's no real precedent at all. Your method handles the first beautifully "when analogous circumstances come along, you know what to do." But in my experience of 23 years in upstream oil and gas, the decisions that mattered most were the ones with no analogue. Frontier basins, sudden geopolitical ruptures. Nothing to back-test against. Written principles still helped, but not by telling me the move, they helped by making clear which assumptions I was betting on.
So, a gentle push on "timeless and universal": back-testing across every era and country still only samples the one history that actually happened. Which means the human in the partnership has two jobs, not one. Aligning with the system inside the world it knows and noticing the moment you've stepped outside that world. The system can't raise that flag itself; by construction it doesn't know where its world ends, especially when it is pulled in opposing direction by many forces and constraints. My worry with deep AI partnership is that the first skill gets stronger while the second quietly atrophies. The sentry at the edge still has to be human.