Dreams are the default for intelligence
kk.org
kk.org
The idea seemed to be: if you could get your dreaming mind to look at your hand and then turn it over to the other side, you didn't have the mental 'bandwidth' to imagine the opposite side correctly, and you'd realize you were dreaming.
Equally, if you were around a bicycle in a dream, and tried to look at the gearing mechanisms etc., it just wasn't possible for your brain to generate that level of detail; it would just change the bicycle into an elephant or what-have-you.
I never got to lucid dreaming, but did notice a similar thing happening. So, I always found it interesting how the mind might switch from internal 'concepts' to external 'reality' in a way that isn't readily available to 'conscious' thought.
Not really going anywhere with this, but if 'AI' can generate better bicycles than our dreaming minds, then…
I've done this myself once or twice in a dream and I actually was able to tell I was dreaming but the shock of my realization woke me up.
Yeah, I've done that a few times and it's been reliable for me.
One time I remember, I was absolutely sure I was awake, so I did the "reading text" test, I read something, looked at something else, then read the thing again, they went "Wait, those aren't the same words. Ah, crap."
Just a little tangent!
I have even had some cases where it did occur to me, and I started lucid dreaming (which was very cool), but then I reverted to my default this-must-be-real dreaming state. That was very weird.
For example I had this dream where I was the guy who got killed in the last bit of The Ring (they think things are OK but Sadako comes for him anyway) and then I was able to lucid dream, control it and beat her up.
https://medium.com/luminasticity/beating-up-sadako-82c5fb3f0...
One thing I find fascinating about dreaming is that it seems that the dream wants to prevent you from staying lucid. When I become aware that I am dreaming I “wake up” and start going about my day, only to realise I am still dreaming, at which point I “wake up” again, and this cycle continues. This is a perceptual barrier that can be broken through, at which point you instantly gain hyper awareness awareness and control over you dream.
What we consider waking life, reality, etc. is the same. The mechanism of the hypnosis is the constant identification with phenomena, which creates the sense of being a person, which is nothing more than a conceptual abstraction.
Whenever I would come semi-aware when I was dreaming I would do this reality check and realize I was lucid and could do what I want. The interesting part to me is that blowing air through my nose and hand twigged me to the fact that I was dreaming, but the technicolor mega-gorilla screaming at the sky and throwing laser bolts did not.
I'd disagree slightly about how awake you want to get though. Obviously don't turn on any lights or stand up. But force yourself to be awake for enough time to get your bearings or you are going to lose consciousness too early.
For some reason this has resulted in persistent lucid dreaming before waking up, daily, for over a month now. The capsules rarely caused this, but something about the tea seems to be affecting the bio-availability, or maybe the powder is just different.
Whatever the root cause, it's been quite fun. But a bit distracting as it makes completely waking up difficult, now that there's this incredible interactive mind-movie I'd rather continue playing vs. start my day...
It does seem like a state of mind worth exploring with some potentially huge utility if mastered. I can't produce such immersive and vivid visuals to navigate mental models in my regular conscious thinking. I'd much rather do the equivalent of lucid dreaming to navigate things like interplaying algorithms and data structures for example.
So... you either don't know what you're taking, or don't want to say in a public forum?
Look at the time or look at letters. They are all fractal gibberish if you look at them closely during dreaming.
But that might just be part of the dream.
I’ve more than once floated into letters that would reveal more and more shapes just like zooming animations of fractals. But it hasn’t caused me to realize it was a dream, let alone take control of it.
Lots of crazy and spacey things happens in dreams, that you just accept.
I can do it, I first figured it out when I was 6 or 7, but why? Might be better to just get some sleep.
Though it is an interesting thought experiment, which implies that we evolved much farther than necessary for dreaming simply because the environment outside dreams as an energy source is so competitive, to a point we're now able to create different kinds of alternate realities as well. Which as so often leads to questions like the purpose of our existence.
To me the idea that dreams exist in order to let the brain filter out unimportant experiences from the day before - or highlight important ones - seems more plausible, especially considering sleep deprivation can lead to hallucinations and more, I don't think this can be explained purely with other processes like the self-cleaning mechanisms of the brain during sleep.
Just musing here but I wonder if that line of thought could shed light on how non-sapient animals experience consciousness as well.
I.e. just neural noise, following pathway patterns by biological necessity, but shaped into conceptual patterns by our overactive pattern matchers which do a ton of work to arrange input into a limited set of "observations".
If that's true, nothing we ever dream is truly novel, but just assembled or permuted from the giant set of our own known or imagined pieces.
Which seems pretty consistent with my dreams, at least. And it explains the phenomenon be being unable to recognize the dream state without altering it. Heisendreams!
I have a pet theory that holds that primarily attentive ADHD (ADHD-PI) is actually a facet of the type of autism that used to be called Asperger's.
And that ADHD, generally, is an extension just outside of the autism spectrum: possibly bridging into it with ADHD-PI if not the other ADHD classifications.
One of ADHD-PI's most noticeable traits is virtually uncontrollable daydreaming. When auditory / cognitive fatigue or fast-arriving boredom sets in, for example.
Asperger's theory is plagued by glut. Which makes the condition challenging to accurately comprehend, but at the same time doesn't necessarily imply that a lot of the theory is incorrect. Instead, it may imply complexity. Which makes sense given that this condition isn't acquired but instead is characterized by divergent brain architecture and talent / cognitive resource re-parsing from "normal". Similar in concept to an RPG game.
One facet of this glut of Asperger's theory is the school of thought that hypothesizes that Asperger's is some kind of evolutionary forward expression of a type of intelligence. Which is both generous and an unusually positive take within the glut of theory, ignoring for a moment any negatives.
My hypothesis adjusts this a bit, and proposes that Asperger's may be a type of genetically de-selected (low reproduction chance) contribution of a specific intelligence type. Which may have existed in human populations from the beginning. Implying that Aspergian intelligence, which may be associated with an increase rate in daydreaming, isn't so much a forward evolution but a population mainstay that may serve some type of function in the sense of low-rate genetic contribution.
Say having to do with visual processing that implies advanced talent in pattern processing specifically.
Some human intelligence academics, who are not involved in Asperger's research, place emphasis on pattern processing as the possible heart of it.
I can think of obvious reasons as to why whatever group genetic "intelligence" governs genetic evolution, as well as the roughly equal reproduction ratio of men and women as another example, would "want" to bias human evolution as much as possible away from a high rate of human-specific-intelligence gain. Possibly smoothing it out away from purely Aspergian expression and blending it, over time, with our other natures and talents.
Well, autism is said to related to percent of Neanderthal. e.g. (2016) https://newsroom.uw.edu/story/human-neanderthal-gene-varianc...
You don't need to see things with your inner eye to be able to conjure them in your fantasies. Daydreaming is just being preoccupied on something that only exists in their mind. It's not a strictly visual phenomenon.
They state that generally females are diagnosed adhd-pi higher, but there may be other factors. However, Females are generally diagnosed less with autism:
(From wikipedia) "Several theories exist to explain the sex-based discrepancy, such as a genetic protective effect,[7][8][9] the extreme male brain theory[10][11] and phenotypic differences in the presentation between sexes,[9][12][13][14] which may all be intertwined. Researchers have also debated whether a diagnostic gender bias has played a role in females being underdiagnosed with autism spectrum disorder.[15] Researchers have also speculated a gender bias in parental reporting due to the expectations and socialization of gender roles in society.[16]"
Thought I would indulge your theory, and give more food for thought.
This can really only be partially true for lucid dreaming, though, right?
Maybe all dreaming is really just lucid dreaming. We're all usually just in some kind of split-brain patient situation where around half our brain is "lucidly" controlling a dream and the other half just goes along for the ride.
A lot of dreams are just kinda nonsense, but some I've had are pretty clearly well-thought-out metaphors for things that are happening in my life. Somebody or something pretty smart that knows a lot about me had to have come up with it.
I've had a lot of lucid dreams over the years (starting when I first was prescribed medication to help me sleep when I was a kid, where I'd have them several times a week, and gradually becoming less common over the years to maybe once a month). At least in my experience, being aware that I'm dreaming gives me the ability to choose how to act, the ability to force myself to wake up, and occasionally the ability to break minor "physics" rules (I pretty commonly find in lucid dreams that I can hover a few inches above ground and float around that way instead of walking), but I've never experienced the ability to change my environment, and the "people" in my dreams I interact with continue to act independently.
Obviously my dreams are entirely a creation of my own mind, so some part of my brain is in control of it, but being aware of that isn't sufficient to let me control it arbitrarily any more than I can stop myself from feeling physical pain or disliking the taste of some food in real life. Ultimately, we're not really in "full" control of how our mind and body react to real-life stimulus either; if we were, things like depression and phobias wouldn't exist.
https://news.ycombinator.com/item?id=35516757
And what probably gives us awareness or consciousness is the multiple GPT like structures running in parallel. It really does make me wonder if we are really close to true AGI. It does feel like it.
Very exciting times! Just hope we don't end up doing a Terminator.
Still, even though we've been saying it for 30 years, exciting times indeed! #skynet
There are only 20 Billion Neurons in the Neocortex. GPT-3 has 175 Billion. GPT-4 scores 130 on IQ tests. They are absolutely in the realm of the power of human brains. They're just exposed to different stimuli and have very narrow input/output.
It's more accurate to look at an organic neuron as an entire computing device than a simple transistor.
Then there must be something really wrong with IQ tests. GPT-4 to me seems more like an idiot savant.
No, parameters are loosely analogous to synapses (they are weights on the connections between neurons) not neurons, and the neocortex is not the whole of the brain. There are about 140 trillion synapses in the neocortex, 600 trillion in the brain. And neural connections may not explain the brain's capacity, as mutual interaction with other body systems is involved.
> GPT-4 scores 130 on IQ tests.
Methodological issues (or simply methodological unclarity) with the reports of GPT-4 test results aside, IQ tests do not directly test general intelligence, even under the assumption that g is the same thing as general intelligence. They are a proxy which works for testing that in humans, because in humans the capacities directly tested by them correlate with g. We have no idea what IQ testing results of LLMs, even conducted appropriately, mean, if anything.
> They are absolutely in the realm of the power of human brains.
No, to steal from Pulp Fiction, an LLM and a human brain “ain’t even the same f###ing thing. […] Ain’t no f###ing ballpark neither. […] It ain’t the same league. It ain’t even the same f###ing sport.”
As a side note, it really bothers me when dreaming is used as a plot device in books. I can’t relate at all to the concept of dreaming and remembering the dream.
So I could see how one of these types of devices could be handy to gently bring the consciousness out of deep sleep.
> Go back a half-century, and television’s impact on our closed-eye experiences becomes even clearer. In the 1940s, studies showed that three-quarters of Americans, including college students, reported “rarely” or “never” seeing any color in their dreams. Now, those numbers are reversed.
If you have some combination of a life of habit and adequate rest, you'll less often fool yourself into thinking you dreamed. If you immediately start writing when you wake up, you'll fool yourself into thinking you have the most elaborate and profound dreams ever, although you'll code this as remembering.
Dreams are so insubstantial that we very recently believed that everybody dreamed in black and white, and that dreaming in color was a sign of mental illness. This is a belief that faded completely and quickly alongside the rise of color tv and film. The implications of this are pretty devastating. People themselves believed they only dreamed in black and white, and later largely forgot that this was their experience.
https://www.nytimes.com/2008/12/02/health/02real.html
Why Did We Think We Dreamed in Black and White?
http://www.faculty.ucr.edu/~eschwitz/SchwitzAbs/DreamB&W.htm
Media Influences How We Recall Our Dreams -- Do We Dream In Black And White Or Technicolor?
https://www.sciencedaily.com/releases/2002/12/021224091445.h...
-----
edit: so I disagree with the OP. There are no dreams, only recollections of dreams. When you look at the "dreamlike" imagery of generative AI, you're looking at something similar to what awake people construct dreams from, not dreams themselves.
Probably not healthy for you though!
Remember how Stable Diffusion works? It's essentially noise reduction cranked up to eleven. You train a neural network to clean up slightly noisy images. When it gets good at it you turn up the noise a notch and repeat. You keep doing that until the network will look at pure noise and "repair" it into an image of whatever.
Animal brains have to do something similar, because our senses suck. Take our eyes; the part of your field of view which is actually in focus is tiny [1], so you are constantly updating an internal 3D model of the world, one small patch at a time [2], interpolating between patches and editing out things like your nose to replace it with a hallucination of whatever it happens to be occluding (except right now, because I reminded you of it).
In this sense, what you "see" is not so much the real world as a hallucination guided by your sensory input.
What happens when you reduce the information content of that sensory input, either by making it "unstructured" [3] or by removing it altogether [4]? Same thing that happens to Stable Diffusion when you present it with pure noise: it hallucinates freely.
So it makes sense that lying down on a comfortable bed in a dark, quiet room and shutting your eyes - all steps toward sensory deprivation - should have a similar effect.
[1] https://en.wikipedia.org/wiki/Vision_span
[2] https://en.wikipedia.org/wiki/Saccade