I’ve been meditating on the broader relationship of this post, and Guzey’s, to the idea of instrumental rationality and the question of what rationality is good for. A idea is that instrumental achievement is bottlenecked by akrasia.
When we study sleep, we’re really studying one aspect of fatigue, which is similar if not identical to akrasia. We can think of fatigue as being subdivided into physical and experiential components. Like other forms of fatigue, sleep pressure manifests both in physical brain structures and in our conscious, felt experience.
Sleep may be a specific solution to a specific physical or felt need, and our sense of sleep pressure may be well-tuned by evolution to indicate the amount of sleep required for optimal function. Sleep may also be a nonspecific solution to these needs, and sleep pressure may be only marginally attuned to the timing and duration of required sleep.
More broadly, we have agentic goals. To accomplish them, we generally must maintain our minds and bodies in order to work efficiently and maintain motivation, while also freeing up time to do that work.
It is possible to carefully calibrate many of our routines. For our diets, we can count calories, ensure our daily intake of vital nutrients, prepare delicious or easy-to-cook meals. For exercise, we have access to a wide variety of exercise equipment and clothing, have several rich theories of how to build strength, flexibility, and endurance, and tools to monitor and plan our workouts.
Yet for managing fatigue, our theories, tools, and equipment often seem far less robust, precise, and thoughtful. Activities that seem relevant to fatigue management, including sleep, meditation, passive entertainment, and taking breaks from work, are often thought of as entirely different spheres of human activity (i.e. sleep, spirituality/therapy, and entertainment). It is not considered unusual, and is indeed considered admirable, to invest a great deal of time, money, and care into developing a healthy and well-considered diet and exercise routine. Why is fatigue management not thought of in the same way?
Fatigue management seems like a concept where rationalists have an opportunity to make real progress. What’s needed is a set of terms and frameworks for naming, relating, and making predictions about how our behaviors, experience of fatigue, and capacity for productive work interrelate.
For example, I noticed today a drop in my energy level for several hours in the middle of my day. I felt low energy, low motivation, and was making many mistakes in my math homework. I had a lot of work still to do, and hadn’t been up and working for all that long. In addition, I was worried that if I stopped, that I would get sucked into leisure activities for the rest of the day. After trying to push on and continuing to make little progress, I stopped.
Instead, I watched a couple episodes of TV, went to the grocery store, and cooked myself dinner. When I was finished, several hours later, I found that my head had cleared, and I was once again able to make progress.
Without a considered relationship with fatigue management, the decisions to stop or continue work get made in an impulsive fashion, and do not result in an increased understanding of how my behaviors intersect with fatigue. Because I did engage in conscious fatigue management today, I got a piece of data about how significant fatigue feels, the effect it has on my in-the-moment productivity, and an upper bound on how long it takes to treat that fatigue with food and rest.
Learning to notice fatigue-related performance loss, and to identify which behaviors seem to most efficiently cause or treat fatigue, seems like it could result in a realistic strategy for reducing akrasia over time, being more productive, feeling better, and knowing what to do in order to have better control over one’s mind and body.
I’ve been meditating on the broader relationship of this post, and Guzey’s, to the idea of instrumental rationality and the question of what rationality is good for. A idea is that instrumental achievement is bottlenecked by akrasia.
When we study sleep, we’re really studying one aspect of fatigue, which is similar if not identical to akrasia. We can think of fatigue as being subdivided into physical and experiential components. Like other forms of fatigue, sleep pressure manifests both in physical brain structures and in our conscious, felt experience.
Sleep may be a specific solution to a specific physical or felt need, and our sense of sleep pressure may be well-tuned by evolution to indicate the amount of sleep required for optimal function. Sleep may also be a nonspecific solution to these needs, and sleep pressure may be only marginally attuned to the timing and duration of required sleep.
More broadly, we have agentic goals. To accomplish them, we generally must maintain our minds and bodies in order to work efficiently and maintain motivation, while also freeing up time to do that work.
It is possible to carefully calibrate many of our routines. For our diets, we can count calories, ensure our daily intake of vital nutrients, prepare delicious or easy-to-cook meals. For exercise, we have access to a wide variety of exercise equipment and clothing, have several rich theories of how to build strength, flexibility, and endurance, and tools to monitor and plan our workouts.
Yet for managing fatigue, our theories, tools, and equipment often seem far less robust, precise, and thoughtful. Activities that seem relevant to fatigue management, including sleep, meditation, passive entertainment, and taking breaks from work, are often thought of as entirely different spheres of human activity (i.e. sleep, spirituality/therapy, and entertainment). It is not considered unusual, and is indeed considered admirable, to invest a great deal of time, money, and care into developing a healthy and well-considered diet and exercise routine. Why is fatigue management not thought of in the same way?
Fatigue management seems like a concept where rationalists have an opportunity to make real progress. What’s needed is a set of terms and frameworks for naming, relating, and making predictions about how our behaviors, experience of fatigue, and capacity for productive work interrelate.
For example, I noticed today a drop in my energy level for several hours in the middle of my day. I felt low energy, low motivation, and was making many mistakes in my math homework. I had a lot of work still to do, and hadn’t been up and working for all that long. In addition, I was worried that if I stopped, that I would get sucked into leisure activities for the rest of the day. After trying to push on and continuing to make little progress, I stopped.
Instead, I watched a couple episodes of TV, went to the grocery store, and cooked myself dinner. When I was finished, several hours later, I found that my head had cleared, and I was once again able to make progress.
Without a considered relationship with fatigue management, the decisions to stop or continue work get made in an impulsive fashion, and do not result in an increased understanding of how my behaviors intersect with fatigue. Because I did engage in conscious fatigue management today, I got a piece of data about how significant fatigue feels, the effect it has on my in-the-moment productivity, and an upper bound on how long it takes to treat that fatigue with food and rest.
Learning to notice fatigue-related performance loss, and to identify which behaviors seem to most efficiently cause or treat fatigue, seems like it could result in a realistic strategy for reducing akrasia over time, being more productive, feeling better, and knowing what to do in order to have better control over one’s mind and body.