As Svyatoslav suggests, the most effective way to approach this is to jointly explore with your son what is actually true about various drugs and their benefits and costs, and why they’re treated the way they are. These are not necessarily easy questions to answer and could take some effortful research to actually get to the bottom of things. (I particularly like the advice here.) I doubt 3hr is enough.
I think one of the problems with his argument is that the anecdata he’s collected from blogs and forums is weighted too highly in comparison to actual studies. He says he skimmed some papers that came up in Elicit searches but I don’t know if that’s enough, as they don’t seem to have left him with a sufficient idea of the risks (or he’s just really risk tolerant, as teenagers tend to be).
He also argues that everyone is different and the only way to find out if certain drugs will be beneficial or harmful is by experimenting on yourself, so “why keep reading studies” past a point, but this seems flawed especially without a concrete idea of the possible benefits.
He’s probably seen this post by now, and I’m sure it’ll result in a conversation that leads to a stronger conclusion, so I appreciate the contributions from everyone!
This is essentially a problem of prediction and forecasting, and you might both benefit from looking into the literature on super forecasting. Phillip Tetlock is the go-to, and I’d also look into Nicholas Nassim Taleb.
Basically, your son is right that when you take drugs, you’re placing a bet on the outcome, and the only way to find out for sure what your outcome will be is to take the drug.
However, he seems to be neglecting the critical point that any smart forecaster works very hard to establish the base rate before they make their prediction. That is what you read the research studies for. Of course if he’s just looking for ammunition to justify taking drugs, there really isn’t much of a point in doing more research. But if he’s at least giving lip service to the idea of making smart, well-thought-out choices without getting stuck in analysis paralysis, there are ways to go about the reasoning under uncertainty aspect of the problem. I’d recommend starting with Superforecasting and maybe The Black Swan, and then circle back.
Thanks!
I think one of the problems with his argument is that the anecdata he’s collected from blogs and forums is weighted too highly in comparison to actual studies. He says he skimmed some papers that came up in Elicit searches but I don’t know if that’s enough, as they don’t seem to have left him with a sufficient idea of the risks (or he’s just really risk tolerant, as teenagers tend to be).
He also argues that everyone is different and the only way to find out if certain drugs will be beneficial or harmful is by experimenting on yourself, so “why keep reading studies” past a point, but this seems flawed especially without a concrete idea of the possible benefits.
He’s probably seen this post by now, and I’m sure it’ll result in a conversation that leads to a stronger conclusion, so I appreciate the contributions from everyone!
This is essentially a problem of prediction and forecasting, and you might both benefit from looking into the literature on super forecasting. Phillip Tetlock is the go-to, and I’d also look into Nicholas Nassim Taleb.
Basically, your son is right that when you take drugs, you’re placing a bet on the outcome, and the only way to find out for sure what your outcome will be is to take the drug.
However, he seems to be neglecting the critical point that any smart forecaster works very hard to establish the base rate before they make their prediction. That is what you read the research studies for. Of course if he’s just looking for ammunition to justify taking drugs, there really isn’t much of a point in doing more research. But if he’s at least giving lip service to the idea of making smart, well-thought-out choices without getting stuck in analysis paralysis, there are ways to go about the reasoning under uncertainty aspect of the problem. I’d recommend starting with Superforecasting and maybe The Black Swan, and then circle back.