Utilizing TDT gave me several key abilities that I previously lacked. The most important was realizing that what I chose now would be the same choice I would make at other times under the same circumstances.
This is similar to the mind hack I am working on to bypass my own hyperbolic discounting.
I assume that I will always make the same choice in similar circumstances. I find that this is a very good approximation of my actual behavior.
I determine the potential consequences of the alternatives in relation to my goals. Sometimes it helps me if I specify the consequences in a way that captures an opportunity cost. For example instead of cost in dollars, I’ll consider the cost in terms of new tires for my truck.
I decide what to do—treating the consequences as though they will occur immediately. In practice I only focus on the top one or two consequences for each alternative—based on my current value weighting.
For example, every morning at work I am tempted by the pile of donuts in my office’s cafeteria.
If I ate a donut every day, in a year I could gain an extra 13 pounds (50 work weeks 5 days per week 180 calories per donut / 3500 calories per pound).
These donuts would cost me about $190 (50 work weeks 5 days per week 0.75 dollars per donut).
I could consider more consequences, but these are enough. I don’t want to pay $190 and gain 13 lbs of weight today—just for the enjoyment of the donuts. In fact I would probably pay $190 just to lose 13 lbs right now; forget the donuts.
When I started to implement this approach I discovered that I could engage in automatic behavior that was counter to my choice. For example I would choose to not buy the donut only to get up and walk to the cafeteria. I would reaffirm the choice and yet still select a donut and pay for it. This behavior had almost an alien hand sense to it.
To break those automatic behaviors I found that I could simply stop and refuse to do anything that wasn’t in keeping with my intellectual choice; I would even close my eyes. Every time I had an “urge” and would start to do or to think something, I would stop and check it against my current goal; if it didn’t match I would I would refuse to continue. With repetition this replaced the negative automatic behavior with positive behavior.
It seems to me that one needs to place a large amount of trust in one’s future self to implement such a strategy. It also requires that you be able to predict your future self’s utility function. If you have a difficult time predicting what you will want and how you will feel, it becomes difficult to calculate the utility of any given precomittment. For example, I would be unconvinced that deciding to eat a donut now means that I will eat a donut every day and that not eating a donut now means I will not eat a donut every day. Knowing that I want a donut now and will be satisfied with that seems like an immediate win, while I do not know that I will be fat later. To me this seems like trading a definite win for a definite loss + potential bigger win. Also, it is not clear that there wouldn’t be other effects. Not eating the donut now might make me dissatisfied and want to eat twice as much later in the day to compensate. If I knew exactly what the effects of action EAT DONUT vs NOT EAT DONUT were (including mental duress, alternative pitfalls to avoid, etc), then I would be better able to pick a strategy. The more predictable you are, the more you can plan a strategy that makes sense in the long term. In the absence of this information, most of just ‘wing it’ and do what seems best at the given moment. It would seem that deciding to be a TDT agent is deciding to always be predictable in certain ways. But that also requires trusting that future you will want to stick to that decision.
This is exactly right, which is why I suggest documenting how you respond to different behaviors. I think that it’s only partly deciding to be predictable; it’s also noticing in which ways you already ARE predictable. In a lot of aspects of life, there are patterns in your behavior, you just haven’t noticed them yet.
I pretty much know how little I can eat before it becomes unsustainable/distracting. This is the advantage of actually keeping a record. (I might be able to push below that threshold but at the moment it doesn’t seem worthwhile.) I also have noticed that I eat better when constrained by rules than when trying to follow “good judgment.” EAT DONUT, in particular, is bad for me.
I’ve also made observations about how I feel on different amounts of sleep, and how many hours of work I can maintain before going crazy. (It’s much easier for me to “push” on my work capacity than to “push” on food past a certain point.)
In other words: it’s worth it to try to know yourself better so that you know what “EAT DONUT” will do to you.
What methods do you usually use to keep track of stuff like this? I can well believe that documenting my calories, sleep, exercise, work, etc. would more than pay for itself in terms of hours and dollars, but I’m a little concerned about the upfront willpower/morale investment...I’d feel a little crazy keeping such rigorous track of what I’m doing, and it would help me overcome that feeling if I knew, in detail, what somebody else had done and why/how it worked for her.
I’m especially curious as to whether and how you make any effort to measure the quality of your hours/calories...if you work an hour while distracted, do you count that the same as if you were focused? If you sleep an hour from 10 am to 11 am, do you count that the same as if you sleep an hour from 2 am to 3 am? And so on.
I’m less quantified than I could be, but enough that it makes a big difference over “nothing.” I log hours on Joe’s Goals; I have to count distracted hours the same as focused ones (what else would I do?) but I make up for it by also counting tasks completed.
Sleep I don’t log regularly, except that now I notice what time I go to bed (I didn’t before!) so that I’m actually aware of how many hours I sleep a night.
If you have a difficult time predicting what you will want and how you will feel, it becomes difficult to calculate the utility of any given precomittment.
I worry about making precommitments for many of the reasons you bring up; our natural tendency toward hyperbolic discounting makes sense when we need to reason in the face of uncertain risks. But I have found that when I focus on long term uncertainty I tend to lock myself into my current behavior even if it works against my current goals.
It would seem that deciding to be a TDT agent is deciding to always be predictable in certain ways. But that also requires trusting that future you will want to stick to that decision.
To avoid the uncertainty inherent in making a commitment, my approach is about how to make a choice for right now—based on my current goals. By choosing to not eat a donut right now, I am not deciding anything about my behavior tomorrow. Tomorrow I may have to repeat the same process of reasoning; if my state tomorrow is similar to my state today I will probably make the same choice, but if it isn’t similar I may make a different choice. No guilt, no fuss.
I am using my assumption—that I will always make the same choice in similar circumstances—to help scope and quantify the consequences of my alternatives. In the case of my example it allows me to scale the consequences to a level I can more easily compare to my goals. In a year I want to weigh 10 lbs less than now so eating 13 lbs of calories as donuts appears to work against that goal.
This approach allows me to make an immediate decision which supports my long term goals, while only experiencing the actual risk of this specific choice, and not the combined risk of all similar future choices. If I discover unexpected negative consequences from this choice then my state will be changed; which I will take into account the next time I face similar circumstances. For example if I discover that not eating a 180 calorie donut in the morning leads to me eating 300 additional lunch and dinner calories, then clearly I will start choosing to eat donuts in the morning. But in fact I discovered that the opposite was generally true; when I ate a donut in the morning I tended to eat 200-300 more calories during the rest of the day.
If I knew exactly what the effects of action EAT DONUT vs NOT EAT DONUT were (including mental duress, alternative pitfalls to avoid, etc), then I would be better able to pick a strategy.
By lowering my perceived exposure to risk I felt free to experiment; this allowed me to collect the knowledge I needed to make the best choice for me, for now. I found that weight gain was a real consequence of eating donuts, and not just because of donut calories. I found that I would start to feel groggy a couple hours after eating a donut. I found that I couldn’t just eat donuts occasionally; if I ate donuts on some mornings I would begin to crave them until I ate them every morning. These cravings would last for days after I stopped eating them—often triggered by context and not just hunger—indicating some form of addiction. The biggest negative consequence to not eating donuts was that I had to manage the daily temptation and the associated automatic behavior. At first I had to repeat my reasoning multiple times a day; but now I only have to do it occasionally—there are many days that I don’t even notice or think about the donuts.
I suspect that some people behave more predictably and/or can predict their own behaviour better than others (I don’t think those two things are the same, or necessarily correlated). Which would make it easier to be a TDT agent. Mood stability might be a factor.
This is similar to the mind hack I am working on to bypass my own hyperbolic discounting.
I assume that I will always make the same choice in similar circumstances. I find that this is a very good approximation of my actual behavior.
I determine the potential consequences of the alternatives in relation to my goals. Sometimes it helps me if I specify the consequences in a way that captures an opportunity cost. For example instead of cost in dollars, I’ll consider the cost in terms of new tires for my truck.
I decide what to do—treating the consequences as though they will occur immediately. In practice I only focus on the top one or two consequences for each alternative—based on my current value weighting.
For example, every morning at work I am tempted by the pile of donuts in my office’s cafeteria.
If I ate a donut every day, in a year I could gain an extra 13 pounds (50 work weeks 5 days per week 180 calories per donut / 3500 calories per pound).
These donuts would cost me about $190 (50 work weeks 5 days per week 0.75 dollars per donut).
I could consider more consequences, but these are enough. I don’t want to pay $190 and gain 13 lbs of weight today—just for the enjoyment of the donuts. In fact I would probably pay $190 just to lose 13 lbs right now; forget the donuts.
When I started to implement this approach I discovered that I could engage in automatic behavior that was counter to my choice. For example I would choose to not buy the donut only to get up and walk to the cafeteria. I would reaffirm the choice and yet still select a donut and pay for it. This behavior had almost an alien hand sense to it.
To break those automatic behaviors I found that I could simply stop and refuse to do anything that wasn’t in keeping with my intellectual choice; I would even close my eyes. Every time I had an “urge” and would start to do or to think something, I would stop and check it against my current goal; if it didn’t match I would I would refuse to continue. With repetition this replaced the negative automatic behavior with positive behavior.
It seems to me that one needs to place a large amount of trust in one’s future self to implement such a strategy. It also requires that you be able to predict your future self’s utility function. If you have a difficult time predicting what you will want and how you will feel, it becomes difficult to calculate the utility of any given precomittment. For example, I would be unconvinced that deciding to eat a donut now means that I will eat a donut every day and that not eating a donut now means I will not eat a donut every day. Knowing that I want a donut now and will be satisfied with that seems like an immediate win, while I do not know that I will be fat later. To me this seems like trading a definite win for a definite loss + potential bigger win. Also, it is not clear that there wouldn’t be other effects. Not eating the donut now might make me dissatisfied and want to eat twice as much later in the day to compensate. If I knew exactly what the effects of action EAT DONUT vs NOT EAT DONUT were (including mental duress, alternative pitfalls to avoid, etc), then I would be better able to pick a strategy. The more predictable you are, the more you can plan a strategy that makes sense in the long term. In the absence of this information, most of just ‘wing it’ and do what seems best at the given moment. It would seem that deciding to be a TDT agent is deciding to always be predictable in certain ways. But that also requires trusting that future you will want to stick to that decision.
This is exactly right, which is why I suggest documenting how you respond to different behaviors. I think that it’s only partly deciding to be predictable; it’s also noticing in which ways you already ARE predictable. In a lot of aspects of life, there are patterns in your behavior, you just haven’t noticed them yet.
I pretty much know how little I can eat before it becomes unsustainable/distracting. This is the advantage of actually keeping a record. (I might be able to push below that threshold but at the moment it doesn’t seem worthwhile.) I also have noticed that I eat better when constrained by rules than when trying to follow “good judgment.” EAT DONUT, in particular, is bad for me.
I’ve also made observations about how I feel on different amounts of sleep, and how many hours of work I can maintain before going crazy. (It’s much easier for me to “push” on my work capacity than to “push” on food past a certain point.)
In other words: it’s worth it to try to know yourself better so that you know what “EAT DONUT” will do to you.
What methods do you usually use to keep track of stuff like this? I can well believe that documenting my calories, sleep, exercise, work, etc. would more than pay for itself in terms of hours and dollars, but I’m a little concerned about the upfront willpower/morale investment...I’d feel a little crazy keeping such rigorous track of what I’m doing, and it would help me overcome that feeling if I knew, in detail, what somebody else had done and why/how it worked for her.
I’m especially curious as to whether and how you make any effort to measure the quality of your hours/calories...if you work an hour while distracted, do you count that the same as if you were focused? If you sleep an hour from 10 am to 11 am, do you count that the same as if you sleep an hour from 2 am to 3 am? And so on.
I’m less quantified than I could be, but enough that it makes a big difference over “nothing.” I log hours on Joe’s Goals; I have to count distracted hours the same as focused ones (what else would I do?) but I make up for it by also counting tasks completed.
Sleep I don’t log regularly, except that now I notice what time I go to bed (I didn’t before!) so that I’m actually aware of how many hours I sleep a night.
At least a partial answer is here.
I worry about making precommitments for many of the reasons you bring up; our natural tendency toward hyperbolic discounting makes sense when we need to reason in the face of uncertain risks. But I have found that when I focus on long term uncertainty I tend to lock myself into my current behavior even if it works against my current goals.
To avoid the uncertainty inherent in making a commitment, my approach is about how to make a choice for right now—based on my current goals. By choosing to not eat a donut right now, I am not deciding anything about my behavior tomorrow. Tomorrow I may have to repeat the same process of reasoning; if my state tomorrow is similar to my state today I will probably make the same choice, but if it isn’t similar I may make a different choice. No guilt, no fuss.
I am using my assumption—that I will always make the same choice in similar circumstances—to help scope and quantify the consequences of my alternatives. In the case of my example it allows me to scale the consequences to a level I can more easily compare to my goals. In a year I want to weigh 10 lbs less than now so eating 13 lbs of calories as donuts appears to work against that goal.
This approach allows me to make an immediate decision which supports my long term goals, while only experiencing the actual risk of this specific choice, and not the combined risk of all similar future choices. If I discover unexpected negative consequences from this choice then my state will be changed; which I will take into account the next time I face similar circumstances. For example if I discover that not eating a 180 calorie donut in the morning leads to me eating 300 additional lunch and dinner calories, then clearly I will start choosing to eat donuts in the morning. But in fact I discovered that the opposite was generally true; when I ate a donut in the morning I tended to eat 200-300 more calories during the rest of the day.
By lowering my perceived exposure to risk I felt free to experiment; this allowed me to collect the knowledge I needed to make the best choice for me, for now. I found that weight gain was a real consequence of eating donuts, and not just because of donut calories. I found that I would start to feel groggy a couple hours after eating a donut. I found that I couldn’t just eat donuts occasionally; if I ate donuts on some mornings I would begin to crave them until I ate them every morning. These cravings would last for days after I stopped eating them—often triggered by context and not just hunger—indicating some form of addiction. The biggest negative consequence to not eating donuts was that I had to manage the daily temptation and the associated automatic behavior. At first I had to repeat my reasoning multiple times a day; but now I only have to do it occasionally—there are many days that I don’t even notice or think about the donuts.
I suspect that some people behave more predictably and/or can predict their own behaviour better than others (I don’t think those two things are the same, or necessarily correlated). Which would make it easier to be a TDT agent. Mood stability might be a factor.