5 ms·
Pharmacology of Hallucinations
- 65a 3y agoHallucination is a terrible model for LLM factual errors, perhaps as terrible as I could think of. Most pharmacological true hallucinogens are all about the brain not being able to properly receive input (anticholingergics, e.g. diphenhydramine, BZ, cyclizine, etc) and misinterpreting it with additional neuronal functional degradation due to synaptic interference by neurotransmitters (or lack thereof). This isn't really the problem that LLMs have, which in my opinion is much more akin to classical confabulation: https://en.wikipedia.org/wiki/Confabulation https://en.wikipedia.org/wiki/Confabulation
- kromem 3y agoWhile I don't think the term will switch at this point, you are dead right. From early on it reminded me of the anecdotes where people with a severed corpus callosum would have their left brain suddenly explain why they had just laughed, completely unaware their right brain has just been shown fake poop. Confabulation is a great word for it, and has a similar quality. But I suspect we'll end up having grown past the issue long before we'd see the term switch given the entrenched usage by now.
- B1FF_PSUVM 3y agoThanks, the description as confabulation is helpful, and should be more widely known. Maybe the current usage of hallucination can be viewed as metonymy or perhaps synecdoche - the untrained observer sees absurd outputs, the description of the phenomenon does not include the reason.
- sorokod 3y agoI like confabulation too but ultimately one way or the other we just anthropomorphizing the software. Does anyone seriously considers LLMs as modeling the human brain?
- dr_dshiv 3y agoWith confabulation it isn’t anthropomophizing but rather drawing a parallel to an information processing approach in brains. I don’t like hallucination because it implies that it is different from how it normally works. Theories of confabulation suggest that this normal brain behavior. Still not the best word.
- johnisgood 3y agoTo note, hallucinations (actually, they cause delirium) from anticholinergics are very different from hallucinations from psychedelics, which are closer to visual distortions instead.
- 65a 3y agoIt's arguable that those are not hallucinations, as there is not normally a misunderstanding about their (physical world) reality, just some amount of acknowledged distortion of perception of the real world.
- johnisgood 3y agoYeah, you are right, that is why I said that they are close to visual distortions, rather than hallucinations. To top it up, delirium-induced hallucinations differ from disassociatives (such as ketamine or DXM) as well.
- Danjoe4 3y agoTLDR the brain is complex and while the three main classes of psychosis-inducing drugs have distinct principal mechanisms, there's likely a significant degree of overlap.
- Podgajski 3y agoAnd genetics. Genetics matter. I don’t need drugs to create changes in perception or my visual hallucinations. I need drugs to stop them.
- echelon_musk 3y agoTFA mentions: > hallucinations may also be triggered by at least three different kinds of drugs: psychostimulants (i.e., cocaine or amphetamine), the so-called “dissociative anesthetics” (i.e., phencyclidine (PCP) or ketamine), and psychedelics, (i.e., lysergic diethylamid (LSD) and psilocybin). It's my understanding that Psilocybin and LSD do not produce hallucinations (e.g. seeing something which does not exist. I'm aware of the difficulty of qualifying the previous sentence) versus changing your perception of phenomena. The exceptions to this are that these compounds can trigger an individual to enter a psychotic state in which this then becomes possible. Or that in _extremely_ high doses such as the 'white light' territory of LSD. However the vast majority of those who use these substances do not tend to report, for example, conversations with people whom a sober sitter would state were not in the room (like the effects of dissociatives such as Datura). The article seems to agree with my understanding of this nuance. I have found this distinction to be relatively uncommon in reporting otherwise. Initially it was thought that LSD was psychotomimetic (mimicing the symptoms of those with pre-existing psychotic illnesses) and there were hopes that they could be used for study into these conditions. I understand that this is no longer working scientific thinking. A good documentary on these topics is the BBC's LSD The Beyond Within from 1987 [0]. Personally I have reached a personal belief that entheogens allow you traverse the realms of existence [1] within your lifetime. For example, descending into the hell realm or ascending into the deva realm. [0] https://www.youtube.com/watch?v=8-x_5grftlU https://www.youtube.com/watch?v=8-x_5grftlU [1] https://en.wikipedia.org/wiki/Six_Paths?useskin=vector https://en.wikipedia.org/wiki/Six_Paths?useskin=vector
- justatdotin 3y agoIt's my understanding that Psilocybin and LSD do produce hallucinations, but that these are mere artefacts; a sideshow to the main game.
- B1FF_PSUVM 3y ago> allow you traverse the realms of existence I have the more prosaic belief that using a rickety chair allows you to sit on the floor.
- 3y ago
- clbrmbr 3y ago> There are at least three different pharmacological ways to induce hallucinations: (1) activation of dopamine D2 receptors (D2Rs) with psychostimulants, (2) activation of serotonin 5HT2A receptors (HT2ARs) with psychedelics, and (3) blockage of glutamate NMDA receptors (NMDARs) with dissociative anesthetics. I’m always surprised when the kappa-opioid pathway (salvinorin A) is overlooked. It produces very distinctive hallucinations. I lack the personal experience to judge, but I think it’s phenomenalogically distinct from other dissociate experiences. It should be easier to research given that Salvia diviniorum and its constituents are not controlled federally nor in many states.
- temp0826 3y agoA bit more dangerous recreationally but scopolamine (datura/brugmansia/etc) is hallucinogenic in yet another way (antimuscarinic, blocking acetylcholine receptors).
- tessellated 3y agoAlso, there are Kappa Opioid antagonists like Salvinorin A. Proceed with caution.
- beefman 3y agoSalvinorin A is a k-opioid agonist.
- Wherecombinator 3y agoI’m not familiar with pharmacology of salvia but taking it is quite dangerous compared to other hallucinogens. A number of my friends had terrible feelings and were freaked out for about a week afterwards. Similar stories are found on erowid.
- EMM_386 3y agoYou can also hallucinate severely (visual and auditory) during the delerium tremens (and alcoholic hallucinosis), which are caused by abrupt alcohol withdrawal after heavy and prolonged use. https://en.wikipedia.org/wiki/Delirium_tremens https://en.wikipedia.org/wiki/Delirium_tremens
- yboris 3y agoRelated: using mathematics and neural architecture to explain geometric hallucinations present under LSD: https://plus.maths.org/content/uncoiling-spiral-maths-and-hallucinations https://plus.maths.org/content/uncoiling-spiral-maths-and-ha...
- aatd86 3y agoWhat if the brain can perceive more signals but that the most common signals are those that are "known" to be captured by the average of the population? To put it another way, there could also be cases where what is deemed an hallucination isn't, but it might lead to psychosis if the subject doesn't know how to interpret the new signal. To reason by analogy, that would be akin to everyone being blind and a couple people developping sight. Whole point is that unless we can understand what the brain is doing, we simply observe a phenomenon but we don't know what the phenomenon actually is. (not talking about drug induced hallucinations, just to be clear)
- Ilverin 3y agoThere are plenty of good scientists and engineers who do psychedelics. If what you say were true, I think at least one of them would have figured out what the sense data is caused by in the environment and have set up a demo such that they can reliably react to a signal sent by a skeptical tester who turns the demo device's power switch off and on, in order to provide evidence to skeptics of their superior perception ability.
- mustacheemperor 3y agoRelated and interesting, there are some common visual patterns across many hallucinogenic experiences, “form constants”. Possibly related somehow to the organization of our visual system. It blew my mind when I learned these exist, and especially when I learned there has been apparently little research into them. Maybe more informed minds than I have determined there’s nothing to learn from them, but it seems like there would be insights to gain about how we see. https://en.wikipedia.org/wiki/Form_constant https://en.wikipedia.org/wiki/Form_constant
- a_gnostic 3y agoThe scientific paper: https://www.hindawi.com/journals/bmri/2014/307106/ https://www.hindawi.com/journals/bmri/2014/307106/