Effect of LSD on reinforcement learning in humans (2022) • Hacker News Board

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

#101

Post by robotresearcher » Mon, May 08, 2023, 3:24 PM UTC

GolDDranks wrote: post #98

robotresearcher wrote: post #80

Earlier quoted context omitted.

I read that differently. There are harder problems in NP, in the same way there are 7ft tall people in the set of people over 6ft tall. So ‘at least NP hard’ I understood as ‘in NP but maybe in a subset of NP that is harder’.

Oh, so you mean like, supposing that NPI exists, an NPI problem? https://en.wikipedia.org/wiki/NP-intermediate > There are harder problems in NP That is an unresolved question. It might turn out that every problem in NP is equally hard. (Up to a polynomial translation cost.)

Thanks for the better information!

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

#102

Post by cptcobalt » Mon, May 08, 2023, 4:07 PM UTC

anigbrowl wrote: post #79

yakubin wrote: post #53

Earlier quoted context omitted.

> I'm pretty sure anyone on earth would know whether or not they were on 75 micrograms of LSD I wouldn’t. I have no idea how strong the effects of LSD are supposed to be based on a dose. I suspect only people who have taken LSD know really.

100 micrograms taken orally is the popular norm. What GP means is that you might not know as a reader, but if you were one of the test subjects you would very definitely know you were not in the control group. Likewise if you were in the control group but had taken LSD before you'd know you had drawn the short straw (unless the details of what was being tested were completely withheld).

Withholding the details of what was being tested would definitely not be informed consent.

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

#103

Post by aidenn0 » Wed, May 10, 2023, 4:22 AM UTC

akiselev wrote: post #44

perrylaj wrote: post #40

Earlier quoted context omitted.

I'm not sure how global the practice is, but at least in the US, you're briefed on the study as part of patient consent. They clearly tell you that nature of the study, including details about blinding, placebos that study participants may get instead of the active chemical, etc.

Informed consent was codified into international law [1] in response to Nazi and Japanese human experiments. There's a few notable exceptions like China (signed but not ratified internally) and in some cultures it may not be emphasized enough to doctors in training but it's pretty universally understood in medical ethics that lying to patients is a line we should not cross. [1] https://en.wikipedia.org/wiki/Internati…

That only requires consent, not informed consent. AFAIK, if someone consents to being experimented on, that law does not require you to fully inform them of the experiment, and certainly not to the degree that a typical IRB in the US requires.