Live data from Hacker News

Effect of LSD on reinforcement learning in humans (2022)

cambridge.org

1–10 of 103 posts

Re: Effect of LSD on reinforcement learning in humans (2022)

#2
No expert, but I had some immediate questions when reading this. Wasn't the dose high enough to be easily distinguishable from the placebo without any experience? Why not some other drug a palcebo? Is single blind design actually advisable in such a study? Can someone enlighten me?

Re: Effect of LSD on reinforcement learning in humans (2022)

#4
post #2

No expert, but I had some immediate questions when reading this. Wasn't the dose high enough to be easily distinguishable from the placebo without any experience? Why not some other drug a palcebo? Is single blind design actually advisable in such a study? Can someone enlighten me?

75ug would definitely be 'noticeable' if you took it sublingual. But this was an IV so I wouldnt know how much dilution there would be before passing the blood brain barrier. My best guess is that it would lower effective dose and how strong thr initial onset was.

Re: Effect of LSD on reinforcement learning in humans (2022)

#5
post #2

No expert, but I had some immediate questions when reading this. Wasn't the dose high enough to be easily distinguishable from the placebo without any experience? Why not some other drug a palcebo? Is single blind design actually advisable in such a study? Can someone enlighten me?

> Wasn't the dose high enough to be easily distinguishable from the placebo without any experience?

This was my first thought. There’s no way that anybody would mistake 75 ug of LSD for placebo.

Re: Effect of LSD on reinforcement learning in humans (2022)

#6

Would love to see this type of experiment performed across different dosages, 25 μg -200 μg. I think the "positive" effects would take a big, sudden drop after a certain threshold

Where can I sign up to participate in a study that involves taking 200 ug of LSD? :)

Re: Effect of LSD on reinforcement learning in humans (2022)

#7
Complete BS study. Increased exploration reduces reward learning rate. That’s the whole point of explore-exploit trade off which is at least NP hard.

The way they refer to RL dynamics as “RL rates” is also a give away that these people don’t understand RL. I’ll leave the evaluation of their cognitive science merit to people knowledgeable in that area

Re: Effect of LSD on reinforcement learning in humans (2022)

#8
post #2

No expert, but I had some immediate questions when reading this. Wasn't the dose high enough to be easily distinguishable from the placebo without any experience? Why not some other drug a palcebo? Is single blind design actually advisable in such a study? Can someone enlighten me?

75ug would definitely be 'noticeable' if you took it sublingual. But this was an IV so I wouldnt know how much dilution there would be before passing the blood brain barrier. My best guess is that it would lower effective dose and how strong thr initial onset was.

Almost all drugs show higher potency when administered intravenously...
Post reply on HN