Hacker News new | past | comments | ask | show | jobs | submit login
State-Space of Drug Effects (qualiacomputing.com)
70 points by gwern on July 12, 2016 | hide | past | favorite | 17 comments



The methodology from On Being Stoned is much stronger. The problem is that the qualitative effects of a drug are determined not just by its molecular structure, but also by the dose, the route of administration, and the type of person using the drug.

By not controlling for any of this, you're just seeing the biases of the status quo being reflected back at you.


> The problem is that the qualitative effects of a drug are determined not just by its molecular structure, but also by the dose, the route of administration, and the type of person using the drug.

Environment also plays a role. I took a great deal of LSD for a few years way back when. I got into a rut of having enormously bad experiences for a while and came to believe that there was something terribly wrong with my psychological state, leading me to take it only in solitude so that I could enjoy my favorite music to it without corrupting the experience for others. At some point, I reluctantly agreed to dose with my then gf and was shocked when she told me she had never seen anyone so happy and positive in the midst of a trip.

About a year later, I had resumed taking it with close friends and the bad times seemed long gone. I dosed with another close friend for the first time in two years and had an awful experience. It hit me that he had been with me for every bad trip that I'd ever had. We had been friends for years, but something about our dynamic was causing me tremendous anxiety in the moment. I never took a hallucinogenic drug in his presence again and the bad times again subsided.

On a side note, if you watch some of the army experiments with LSD administered to soldiers in sterile lab rooms from the Cold War era, it's no wonder some of them are suffering so much. It makes me feel bad for them just thinking about it.


you're just seeing the biases of the status quo being reflected back at you

I love your phasing. It captures the essence of many "bugs" in stataistical analyses in a rather poetic and, more importantly, persuasive way.

Is this original or do you happen to have a source for this quote?


I wasn't quoting anything. I'm not a huge stats person, I have a casual interest in epistemology but in this case that sentiment just comes from having a lot of domain-specific knowledge about drugs. If you search my comment history for drug-related stuff, you'll see similar criticisms of David Nutt's studies, among other things.


The best qualitative thing I've seen on the state space of drugs is this selection of self portraits: http://bryanlewissaunders.org/drugs/. Special mentions for meth, lithium, and bath salts. Ooh boy I don't want bath salts.


This is the best resource that I've seen that covers the possible qualitative effects of different drug classes:

http://disregardeverythingisay.com/post/9331287956/the-visua...

The one drug-related thing that's been bothering me the last couple weeks is just how many native psychoactive plants have virtually zero writeups on Erowid: indian tobacco, sweet flag, mugwort, mulberries, indian pipe, etc. At least in New England every single park, even in Manhattan, is pretty much an all-you-can-eat buffet of psychoactive plants and yet there is very little information about the safety and efficacy of most of them, or even a quorum of anecdotes for that matter.


I guess it's sort of a cool art project? But it doesn't seem that useful as a data point. Mostly because the artwork is likely affected by the expected outcome more so than the actual experience, and that experiences can vary wildly even with the same substance, and varies wildly amongst people.

I.e. from this guy's self portrait with absinth you'd imagine that it makes you sad, but that's never been my experience with it.

More than anything it's a nice showcase of the artist's stylistic range.


Agreed, and more generally, how do you say attribute different artistic stylings as due to a drug vs. all other factors (e.g., expectations of the drug, mood, social setting, etc.). I would imagine this has been studied in perceptual psychology (or some related field) in a more controlled fashion.


Wow, that guy is an idiot.


Why?


Look at the incredibly dangerous/unsafe/unfun drugs he did in that list. Duster, meth, huffing naptha... I mean, I'm all for self-experimentation but we know certain things are incredibly damaging to human beings.


Very few drugs are significantly physically dangerous when not used chronically, as long as you're careful with dosage. Even volatile solvent inhalants, easily the least-safe commonly-used drugs, simply can't do that much damage without becoming a habit. The cases of people dying on their first use of inhalants are either overdoses (not hard to overdose on inhalants) or the result of blowing pressurized gas directly into the throat, causing a dramatic drop in temperature.

https://en.wikipedia.org/wiki/Intoxicative_inhalant#Sudden_s...

I imagine that this project had a net-negative effect on his health, but calling him an idiot assumes that he didn't weigh the risks. I get the impression that he has done more research on the topic than we have.


It's odd that the article doesn't even mention Shulgin's PiHKAL and TiHKAL. There are protocols.

Also, without access to accurate testing, people have no clue what compound(s) they're actually using. So I suspect that surveys get mostly at expectations. As Alex3917 notes.


The article itself does not mention it, but the site definitely does, as well as expanding on the works of Lilly, Carhart-Harris, Pearce and a variety of other psychedelicists.

The survey also collected information about the number of times that the substance was used, though apparently it didn't matter. What is remarkable is how well the differences match to common knowledge, even though each survey taker only provided a response for a particular substance. Expectations might have a bigger role if the survey included all drugs at once, and people thought on their own "I don't know about that one, so I'll just answer with what I have heard about it." Rather, the survey asked participants to answer it for a drug that they had substantial experience with. Another thing to note is that factor analysis itself is likely to reduce a lot of possible biases, since people don't know in advance how the various items will end up being bunched together. If there is a popular expectation that ecstasy is "life changing" that particular item might have a weird pattern of responses for ecstasy. And yet, its overall loading on the various factors would remain relatively unaffected, since each item is contributing only a little to the final factor scores. Thus lexicalized associations (e.g. "the beauty of LSD") will be washed away by the influence of the remaining 49 items.

But yes, one wishes a lot of improvements. Personally, though, I think that having random people on the internet applying statistical techniques to new domains of knowledge in order to bring attention to low hanging fruit is really good. Of course they will not do it as well as a funded lab would, but they will help accelerate such research.


Thanks. Perhaps my cynicism about drug authenticity is dated.


It seems like a parsimonious state-space for a drug would just be the agonistic/antagonistic affinity for every neuro-receptor. There would be some partial coronations (between all of the subtypes of the GABA receptor for example).

It would be neat to see how the authors psychological state-space and this neuro state-space map to each other though.


> just be the agonistic/antagonistic affinity for every neuro-receptor

Except that pretty much every receptor subtype, every transporter protein and many other brain genes exist in multiple variations in the human population, and those variations do alter the way these proteins act. This is one reason why different drugs affect different people differently, because at the lowest level, you don't have the same brain chemistry as other people, there are an infinite number of possible variants.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: