The word struggle is defined as 'to make forceful or violent efforts to get free of restraint or constriction'. Sounds unpleasant to me. A decision is 'a conclusion or resolution reached after consideration.' Put them both together and what do you get? A decision struggle is the process of exercising forceful and violent effort in order to come to the right conclusion about how to reach agreement or solve a problem. I read two research papers recently that have me spooked, and thinking a lot about decision struggles.

One is called 'AI, Human Cognition and Knowledge Collapse' by Acemoglu, Kong and Ozdaglar. The other is called 'Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender' by Shaw and Nave. The bottom line across both papers is that people are already outsourcing their reasoning process from a cognitive one (e.g. system 1 intuition and system 2 deliberation) to a third one that Shaw and Nave coined system 3 for 'cognitive surrender' (to AI). The Acemoglu, Kong and Ozdaglar paper finds that when you do that, knowledge in groups of people collapse.

To guard against knowledge collapse, I wanted to come up with a marketing term for the process by which we create new knowledge, and landed on decision struggle. When we struggle with decisions, we come up against the limits of all three systems of decision making that our brains use to achieve outcomes that we are seeking. This struggle, if documented, can be translated into knowledge for the group. If one makes decisions for themselves or for a group of people, and neither struggles with them nor preserves the context created by that struggle, the knowledge of the group will eventually cease to exist.

Every leader of a group of people should be thinking about how to increase the amount of decision struggle happening. As AI automates task work, the work that remains is exercising judgement in various forms. In all of our research into decision science, the most important idea supporting that is that the decision process is organized around a goal. If you organize the decision around a goal, and struggle productively with it, that process is what leads to new knowledge. You can then use that group knowledge to accelerate the process of getting what you want in the future. I've started calling that idea knowledge compounding, the idea that with sufficient decision struggle, you accumulate more useful knowledge over time rather than letting it erode by outsourcing to AI through cognitive surrender.

Let's try simplifying this even more. We all want things. We have motives, we have goals, and we have various ways that we can go about getting them. If we outsource our thinking to AI, there is no way that we can learn in the long run how to use our intuitive and deliberative decision abilities. This impacts our knowledge individually as we surrender it to the AI. And when groups do this, it impacts the ability for the entire group to learn and preserve knowledge. This is basically what happens when everyone uses single player AI agents to do their tasks, while collaborating in shallow ways in places where decisions aren't preserved. It's really the animating force behind what Convictional is building into our product.

Likewise, it's not sufficient just to struggle with decisions. We need to perform that struggle in places where the rest of our group can meaningfully discover the resulting knowledge. The more public we are willing to be, the smarter the group gets and the more able each member can think like the leader does. Transparency and ease of access are necessary in order to improve the context shared by a group. We must struggle in public, which is challenging and goes against certain social norms we all intuitively feel. That said, when we do struggle in public we improve our entire group's ability to achieve its goals.

If we use AI without changing the way that we collaborate, preserve knowledge, struggle with and record decisions, our ability to learn and get the things that we want will perish. May we all aspire to struggle as much as we're able with the decisions in front of us, and preserve that knowledge for others too.