I hereby commit to doing at least 3 of the remaining 4 of these (I don’t know if I’ll have time every week).
I don’t like measuring things by streaks—if you want to do a list I think doing it by total number of challenges completed is better. Streaks are a less accurate indication of effort put in or potential gains achieved and have more potential to create unhealthy incentives.
(I think this instinct comes from something like Noticing the Taste of Lotus, although I’m not really sure how strongly it applies here)
I’ve mulled over this a bit, and think I disagree, and will keep doing streaks.
One of the goals of the challenge is building a culture of practice. I think consistency is an incredibly important part of that. That’s how you get compound returns. A portfolio that grows 7% every year will grow ~30x over fifty years. But a portfolio that grows that much only every other year will only grow about ~5x. (Even though the first one only put in “twice as much effort”.)
I also think many rationalists could benefit a lot from practicing consistency.
Now, if someone does 49 babble challenge but misses one in the middle, sure, it seems annoying for them to fall down the ladder. But maybe we could allow people to miss one week per month, or something, without hurting their score? Similar to the “never miss twice” mindset for habits, which is more important than “never miss”.
I think compound returns is the wrong model as it stands—logarithmic growth seems more appropriate with the current setup. I would expect completing 5 babble challenges to give 80-90% of the benefit of doing 7.
If we practice both babble and prune then the benefits of the two probably do compound somewhat with each other such that doing 2 babble and 2 prune is significantly better than doing 4 of either but this doesn’t really justify streak measuring.
If consistency rather than direct benefit is the target then streaks make some sense. I would say in that case that I would need to be persuaded that this is the correct exercise to learn consistency. At the moment I would categorise it as definitely worthwhile (hence the 3 out of 4 commitment) but not enough to super-prioritise it enough to make a streak-worthy commitment.
I hereby commit to doing at least 3 of the remaining 4 of these (I don’t know if I’ll have time every week).
I don’t like measuring things by streaks—if you want to do a list I think doing it by total number of challenges completed is better. Streaks are a less accurate indication of effort put in or potential gains achieved and have more potential to create unhealthy incentives.
(I think this instinct comes from something like Noticing the Taste of Lotus, although I’m not really sure how strongly it applies here)
Really excited to have you onboard.
I’ve mulled over this a bit, and think I disagree, and will keep doing streaks.
One of the goals of the challenge is building a culture of practice. I think consistency is an incredibly important part of that. That’s how you get compound returns. A portfolio that grows 7% every year will grow ~30x over fifty years. But a portfolio that grows that much only every other year will only grow about ~5x. (Even though the first one only put in “twice as much effort”.)
I also think many rationalists could benefit a lot from practicing consistency.
Now, if someone does 49 babble challenge but misses one in the middle, sure, it seems annoying for them to fall down the ladder. But maybe we could allow people to miss one week per month, or something, without hurting their score? Similar to the “never miss twice” mindset for habits, which is more important than “never miss”.
I think compound returns is the wrong model as it stands—logarithmic growth seems more appropriate with the current setup. I would expect completing 5 babble challenges to give 80-90% of the benefit of doing 7.
If we practice both babble and prune then the benefits of the two probably do compound somewhat with each other such that doing 2 babble and 2 prune is significantly better than doing 4 of either but this doesn’t really justify streak measuring.
If consistency rather than direct benefit is the target then streaks make some sense. I would say in that case that I would need to be persuaded that this is the correct exercise to learn consistency. At the moment I would categorise it as definitely worthwhile (hence the 3 out of 4 commitment) but not enough to super-prioritise it enough to make a streak-worthy commitment.