I deliberately avoided giving a citation because I don’t remember which paper I read that confirmed it, so searching for one that backs up a cached memory to appear more rigorous would be bad epistemic practice.
Instead, my confidence that this is true rests on several pieces of circumstantial evidence:
My experience for it working this way for other drugs.
The SSC survey where the majority of people reported not becoming dependant on other stimulants at therapeutic doses over the long term.
The fact that coffee has become universal to workplace culture (metis knowledge)
The fact that even if coffee gave you a focus boost that nets to 0, being able to borrow energy from the 2/3rd of the day you aren’t working into the 1⁄3 that you are would still boost net productivity.
The fact that I used to believe that I was being clever by never using caffeine because of the idea that there is no free lunch and changed my mind a few years ago.
Other things I can’t recall right now but I know I could recall them if I sat down for several hours trying to remember them. (How could I possibly know this? it happens on a regular basis)
I don’t necessarily expect you to believe it, but it occurred to me that the implicit choice between:
A. showing you the watertight meta-analysis that I’ve spent a week going over with a fine-tooth comb .
B. saying nothing at all and likely leaving you with no responses because everyone else assumes a response needs to do A to be worth giving.
...is one of the reasons why LessWrong is a terrible place to find practical knowledge.
I’d be happy to bet on it being true at at least 4 to 1 odds, although you will have to devise an objective test that can be judged true or false on the original question rather than the proxy question. Then again, even saying I’m willing to bet doesn’t mean much as a bet of $100 still wouldn’t be worth your time to organise on a financial basis. This makes the bet less likely and therefore boosts the credibility of an argument with a costly signal that’s actually far less costly than it appears. (This is currently an open problem.)
A and B aren’t the only choices. You don’t need to show a watertight meta-analysis to state why you believe what you believe and be explicit about your uncertainty.
How do you know?
I deliberately avoided giving a citation because I don’t remember which paper I read that confirmed it, so searching for one that backs up a cached memory to appear more rigorous would be bad epistemic practice.
Instead, my confidence that this is true rests on several pieces of circumstantial evidence:
My experience for it working this way for other drugs.
The SSC survey where the majority of people reported not becoming dependant on other stimulants at therapeutic doses over the long term.
The fact that coffee has become universal to workplace culture (metis knowledge)
The fact that even if coffee gave you a focus boost that nets to 0, being able to borrow energy from the 2/3rd of the day you aren’t working into the 1⁄3 that you are would still boost net productivity.
The fact that I used to believe that I was being clever by never using caffeine because of the idea that there is no free lunch and changed my mind a few years ago.
Other things I can’t recall right now but I know I could recall them if I sat down for several hours trying to remember them. (How could I possibly know this? it happens on a regular basis)
I don’t necessarily expect you to believe it, but it occurred to me that the implicit choice between:
A. showing you the watertight meta-analysis that I’ve spent a week going over with a fine-tooth comb .
B. saying nothing at all and likely leaving you with no responses because everyone else assumes a response needs to do A to be worth giving.
...is one of the reasons why LessWrong is a terrible place to find practical knowledge.
I’d be happy to bet on it being true at at least 4 to 1 odds, although you will have to devise an objective test that can be judged true or false on the original question rather than the proxy question. Then again, even saying I’m willing to bet doesn’t mean much as a bet of $100 still wouldn’t be worth your time to organise on a financial basis. This makes the bet less likely and therefore boosts the credibility of an argument with a costly signal that’s actually far less costly than it appears. (This is currently an open problem.)
You are now making a different claim than your first comment (which was probably false and is definitely is contradicted by papers).
I read him as making additional claims, not just entirely different ones.
A and B aren’t the only choices. You don’t need to show a watertight meta-analysis to state why you believe what you believe and be explicit about your uncertainty.