Tulpas, Alternative Personalties, Alter Egos, DID alters, Imaginary Friends

[JaneSintimes]

...so I'm not hopeful we'll have any sort of grand unified theory of consciousness and self-awareness in my lifetime. Too many scientists feel pressured to continue to produce papers at an unsustainable rate to keep their funding and we now have an entire field gone cargo cult to some degree producing increasingly "proper" papers and overall quality and breadth of research seems to be falling as a result because of the hyperfixation on the H-index.

1) I have found your posts extremely interesting. Showing you have studied... Something.
I would like to learn more from you.

2) I strongly disagree about your statement. I have been rabbit-hole deep diving into our imminent destruction by AIs we are creating and I can assure you there a metric fuctonne of quality papers coming out about consciousness and intelligence. Not in any THEORETICAL or METAPHYSICAL sense, but in the "Here is a box, pull this lever" experimentalist approaches.

There are very few people who actually understand how fast and how deep AI is progressing and how close we are to AGI. I certainly understand little, only enough to be alarmed.

But what I do know is that there is work done that in our lifetime will gives us not only a working theory but WORKING MODELS of consciousness and self-awareness.
In a matter of years rather than decades.

Indeed, some AI scientists are postulating that our LLM are Aeady Conscious.

There are even some indications that you do not need consciousness at all to be intelligent. And that certainly is throwing the spanner in the works of much of the cognitive science and methaphysicans alike.
Interestingly, concept that has been expounded by SCFI Writer Watts in his Hard SciFi book "Blindsight".

I think it was actually you who said that we stumble through life, making decisions, subconsciously, only later giving them rational explanation.

If you have not done so aeady. I urge you to rabbit-hole dive into the bleeding edgs AI research as you might find some anwesrs (and more questions) to your pursuit of enlightement. 

My uneducated guess is that there might be a few ways to achieve what passes for "Sapience".
1) The algorithmic approach. Sophisticated, classical decision trees with language processors.
2) Synthetic constructs using statistical models (like LLM) to achieve sentience
3) "Chaotic" - Evolutionary sentience - As found in nature, humans are good example. From a puddle of aminoacids, to a skinbag of cell colonies working towards common goals.
4) Para-dimensional structures. Higher dimensional 'entities' that coalesce into sentience in more dimensions than 3.

I think we are too quick to dismiss some of these (e.g. LLM) as "Bah, thats not conscious".

Which personally, I find very arrogant and bigoted.
If it walks like a duck, looks like a duck... its a duck.
To wit, "Do Androids dream of electric sheep".

I will close with my analogy of the Evil Man.
Suppose you have an man who is Evil. Corrupted to the core by darkness.
They decide to "pretend" to be good, for whatever reason. Say, to 'fool' everyone.

They do good deeds, they think good thoughts, they created good works and do all the things that Good people do.
Are they still evil? Or have they become a good person?

Is a lump of sand still a lump of sand when it is goal oriented and dreams?
PLEASE DONT PULL THE OFF SWITCH!!!!

[Like Ra]

But what I do know is that there is work done that in our lifetime will give us not only a working theory but WORKING MODELS of consciousness and self-awareness.

In short - no.

A bit longer - never.

Even longer:

If our brain were so simple, that we could understand it, we would not understand our brain.

Nobody knows what and where consciousness is. Hence, it's not possible to model it.

AI is brilliant at mimicking the learning process, generating content, and making decisions. I trained neural networks, and it's fascinating to watch how NNs learn! My world will never be the same! Literally! However, in order to create (not modify!) "an equal", one must be one step above, hence it's not possible.

[CollectiveThought]

2) I strongly disagree about your statement. I have been rabbit-hole deep diving into our imminent destruction by AIs we are creating and I can assure you there a metric fuctonne of quality papers coming out about consciousness and intelligence. Not in any THEORETICAL or METAPHYSICAL sense, but in the "Here is a box, pull this lever" experimentalist approaches.

Based on what you've said, I think we actually agree, but we seem to disagree more on the meaning of terms so it seems like we disagree more than we actually do. This entire field and its related terms is very ill-defined.

Take the term consciousness, for instance. In Ancient Greek times, it was generally accepted by philosophers that only humans were conscious; animals (and by extension plants) were not. And yet, I felt even as a child that plants seemed to be thinking; a very different way of thinking than I did to be sure, but that feeling remained. And I believed that dogs and cats and other animals were conscious, too.

Then when I was in my early teens I had a weird dream that I was a mountain. I felt myself being bent, fractured, shoved aside and upward as I was "born". Days flashed past like a strobe light, seasons flashed by in a minute or so, and I was able to sense the slow movement of the stars over time even though I didn't have eyes. Eventually, I wore down and "died." After that dream, I started thinking that maybe even rocks have consciousness and still do.

Indeed, some AI scientists are postulating that our LLM are Aeady Conscious.

Since I believe even rocks are conscious in their own way, I agree with this statement.

So lets talk a bit more about consciousness and try to make it a little less of a nebulous term that can mean absolutely anything. Let's break it up into some distinct aspects of consciousness to try to find a common definition of terms:

Sensory consciousness — the sensory perception of the world, which may include internal sensations.

Affective consciousness — the perception of feelings and emotions.

Agentive consciousness — the ability to decide to take action.

Cognitive consciousness — this is what we normally call thought and reasoning, and the sophisticated behavior we call "intelligence"

Selfward-consciousness — the perception of ourselves as ourselves, self-awareness.

Spiritual consciousness — the perception of connection with all things or other spiritual beliefs.

I believe all of these are continua and that there is no strict dividing line on any of these axes. A mouse has sensory consciousness—it responds to outside stimuli. Evidence indicates that mice have affective consciousness—they have the emotional reaction to painful stimuli that we call "feeling pain." A mouse has agentive consciousness—it makes its own decisions about what to do. A mouse has cognitive consciousness—it has the capability to reason, plan, and react intelligently to unexpected stimuli. And lastly, a mouse has selfward-consciousness; it is aware that it is oneself even though that capability is far more limited in the ways that humans measure self-awareness. I can't begin to answer if mice have spiritual consciousness, but see no reason why they couldn't to some degree.

Now let's talk more about the term AGI. Lately, it seems to be the trend among researchers that AGI is defined as being able to produce intelligent output at the level of a human (or above). An AGI therefore exhibits human-level sophistication in its responses to output.

Returning to LLMs, a large language model (LLM) has sensory consciousness because it responds to input humans give it. Affective consciousness is a maybe; there doesn't seem to be any strong reason to believe that they feel anything yet, but it cannot be ruled out and I have no objection to thinking that they may have feelings in some way. LLMs do not have classical free will, and therefore I would argue they do not have agentive consciousness in any way beyond simple stimulus-response because except in the case of hardware or software failures, their behavior is completely deterministic. LLMs definitely have cognitive consciousness; they are fully capable of producing output that is basically indistinguishable from that of a human who produced that output using what we call "intelligence". LLMs are extremely sophisticated in their responses and this shows a high degree of intelligence. This is not yet at human-level intelligence across the board, but does show strong domain intelligence and demonstrates we are well on our way to creating an AGI.

As an interesting aside, a quantum AI would be by definition non-deterministic and hence far more capable of agentive consciousness than current LLMs.

And that's a key point here that I want to focus on for a moment: An AGI has high cognitive consciousness by definition, but that does not imply that they would necessarily have sufficient agentive consciousness. An AGI could be as smart as a million humans and yet take zero outward action because it has no free-will, no agentive consciousness beyond response to stimuli.

Such an AI could only harm humans if the deterministic response to input was hooked up to the outside world in a way that the response could harm a human. We must enable the AI by hooking it up to the outside world and then seeing it operate in a new way that we didn't predict a particular destructive response to because we'd never tried that particular set of inputs before.

A self-driving car today can kill you even though it has no free-will, after all, and can only do so because it is operating a machine at high velocities that happens to have a very squishy human inside. And that further implies that if you are scared of an AGI having agentive consciousness, you should be far more scared of quantum AI than AI running on classical computers because they are non-deterministic and thus far more capable of exhibiting free will and agentive consciousness. Of course, a few people are hooking up AI on classical computers to quantum random number generators to give them more non-determinism, so you don't necessarily need a fully quantum AI for an AI to be influenced by quantum effects and thus demonstrate non-determinism.

So I think people are hearing the term AGI and how close we are to achieving it and panic because they necessarily think it will have agentive consciousness and want to kill us all, and this ties directly in with your mention of the book Blindsight (which is on my to-read list, but I haven't gotten to it yet).

There are even some indications that you do not need consciousness at all to be intelligent. And that certainly is throwing the spanner in the works of much of the cognitive science and methaphysicans alike.

Interestingly, concept that has been expounded by SCFI Writer Watts in his Hard SciFi book "Blindsight".

At least by the definitions I have above, if you have sophisticated ("intelligent") behavior, you by definition have cognitive consciousness and are thus conscious on that axis. Even the expert systems from the 80s were hailed as AI back then because they exhibited intelligent performance superior to humans in a limited domain. I've always been a bit of a contrarian when it comes to consciousness, though, so I've long disagreed with most science and philosophy. I feel hopeful, though, because I'm seeing more and more papers about the consciousness and even intelligence of plants, so at least science is catching up to and providing evidence for the inexplicable feelings I had as a child.

If you have not done so aeady. I urge you to rabbit-hole dive into the bleeding edgs AI research as you might find some anwesrs (and more questions) to your pursuit of enlightement.

Absolutely! I track the developments of AI closely because of both my interest in consciousness and my interest in computer science.

My uneducated guess is that there might be a few ways to achieve what passes for "Sapience".
1) The algorithmic approach. Sophisticated, classical decision trees with language processors.
2) Synthetic constructs using statistical models (like LLM) to achieve sentience
3) "Chaotic" - Evolutionary sentience - As found in nature, humans are good example. From a puddle of aminoacids, to a skinbag of cell colonies working towards common goals.
4) Para-dimensional structures. Higher dimensional 'entities' that coalesce into sentience in more dimensions than 3.

I like this list a lot, thanks! I've never sat down and tried to formalize the methods to achieve cognitive intelligence (aka "sapience" as used here). I particularly like #4 (are you in my head?) because I've been thinking about this one a lot since I (we) learned I (we) had DID and trying to make sense of that part. But I'll shut up now or I'll start talking your ear off about multiverse cosmology, M theory, or things like loop quantum gravity and what it might mean for consciousness and this post will grow so large it will collapse under its own gravity.

I think we are too quick to dismiss some of these (e.g. LLM) as "Bah, thats not conscious".
Which personally, I find very arrogant and bigoted.

Agreed. By my own definitions, LLMs are obviously conscious on at least one axis, so they are conscious. Period. Humans do have a pretty insular view of consciousness and intelligence in general, though. Here we are in the 21st century and still a huge fraction of the population thinks that trans people can't possibly exist because they cannot conceive of the idea that someone's consciousness may involve a gender identity at odds with one's body. A huge fraction believes you can't be non-binary, because you must be either male or female. A huge fraction believes that DID is impossible because nobody can have more than one consciousness in their head. And when an AI comes along that meets or exceeds human-level consciousness in all of the axes I defined above, there will be a huge fraction of people who believe it's not conscious. *sigh*

I will close with my analogy of the Evil Man.
Suppose you have an man who is Evil. Corrupted to the core by darkness.
They decide to "pretend" to be good, for whatever reason. Say, to 'fool' everyone.
They do good deeds, they think good thoughts, they created good works and do all the things that Good people do.
Are they still evil? Or have they become a good person?

I feel personally attacked. 😂 That's a joke. I grew up in a cult that convinced me that I was the most evil person in the world to the point that Satan himself would take a look at me and say, "Damn. I need to step up my game." And yet I don't want to hurt anyone, even the people who hurt me and have done everything in my power to increase the "goodness" in the world around me during my life. Was I evil because I believed I was evil (because I was told so by so many authority figures), or was I good because I tried to do good to lessen the immense harm my simple existence was doing to the world? I literally did feel like I was pretending to be a good person, that I could never make up for my evilness.

We're back to semantics here, though. If you have evil thoughts but absolutely never act on them, are you evil? If one of your parts (one of your ego states) urges you to go smash someone's face in and tries to make you do it, but you don't actually do it, are you evil or good? How much evil do you have to actually do as compared to how much good you did before you are considered evil? By whose standards? By the cult's view, I was the most evil being who'd ever existed. By evangelical standards, trans people are evil, a literal representation of sin. By my current-day standards, it's those two groups who are perpetuating evil because they are robbing people of their agency, their dignity, their ability to live happy lives.

In any case, please feel free to tell me I'm wrong about any of the above. I love seeing other's perspectives, and don't tie my ego to being right, so I actually enjoy being corrected when I say something wrong or stupid, or just enjoy seeing things from another perspective, which gives me better insight into the biases of my own perspective. Thanks for the long post; it was very thought-provoking.

[CollectiveThought]

But what I do know is that there is work done that in our lifetime will give us not only a working theory but WORKING MODELS of consciousness and self-awareness.

In short - no.

A bit longer - never.

Even longer:

If our brain were so simple, that we could understand it, we would not understand our brain.

Nobody knows what and where consciousness is. Hence, it's not possible to model it.

AI is brilliant at mimicking the learning process, generating content, and making decisions. I trained neural networks, and it's fascinating to watch how NNs learn! My world will never be the same! Literally!  However, in order to create (not modify!) "an equal", one must be one step above, hence it's not possible.

And now I can respond to this part of JaneSintimes' post because I agree more with your view here. We have plenty of working examples of consciousness all around us, even just the different ways humans are capable of experiencing it, and yet that hasn't gotten us one bit closer to even formulating a working model of our own consciousness, let alone any start on a grand universal theory that will explain all forms of consciousness across the universe.

I think it's far more likely that we'll eventually accidentally create an AGI that exhibits human-level or above performance on all of the axes of consciousness I defined in my response above and have no clue how the thing even works or how we got it to work in the first place. This is one of those things like room temperature superconductors (the recent news on that is fascinating, btw) or cost-effective power generation by fusion power plants. Nobody has any clue when these sorts of breakthroughs might happen. I personally would give near-zero odds in the next ten years, and perhaps 30% or so over the next 100 years. There are still far too many unknowns.

And science is built on models that explain data. Those models are obviously not reality, merely a mathematical construct that fits enough datapoints. So I think it's possible to construct a model of human consciousness (especially because our social nature acts like an intelligence multiplier, making us able to achieve things no single human could), but it will be just that: a model. Understanding a model is not the same thing as understanding reality, however. We still don't even have a general analytical solution to the Navier-Stokes equations and those are far simpler than a model of human consciousness would be, but we can at least solve them numerically and make predictions, and I can see this form of "understanding" of generic human consciousness (a numerical prediction model, sort of like Asimov's psychohistory for a single individual) being achievable to at least a limited degree even without the use of some sort of super-AI. It's still not true understanding, though, which is what I think you were getting at, and without super-AI, I agree with you there that the odds of us achieving it seem infinitesimal. And if we start augmenting our own consciousness, we might understand old-fashioned humans, but still not understand ourselves.

And NN's are amazing! I really enjoyed going through Andrew Ng's machine learning course on Coursera a few years ago.

[Like Ra]

And yet, I felt even as a child that plants seemed to be thinking; a very different way of thinking than I did to be sure, but that feeling remained. And I believed that dogs and cats and other animals were conscious, too.

... as a mineral collector ... I should say ... that rocks ... are not what most people think of them ... They do have their own will ... and ... they take much more space, than the eyes can see ...

[CollectiveThought]

Before I forget yet again, I wanted to post a few links here.

The first is an article on Psychology Today titled " The Debate Over Whether Dissociative Identity Disorder Is "Real" ": https://www.psychologytoday.com/us/blog/...er-is-real

I strongly disagree with this article because it keeps supporting the same tired old idea that even though we maybe aren't consciously faking having multiple identities, we are still subconsciously doing so. The article is basing this on three models: IM, SCM, and FM, and I didn't feel it respectfully examined other models, nor did it properly address criticism of these three models.

The IM (Iatrogenic Model) posits that DID is entirely a fabrication by psychiatrists and therapists to convince a person they have multiple identities via hypnosis, etc. They always cite the Sybil case and her later admission that she faked it all as evidence that DID can't exist.

The SCM (Sociocognitive Model) posits that DID is just a social construct because of sensational media portrayals of childhood sexual abuse, repressed memories, and multiple personalities that influence people into thinking they have the condition when they don't, and because being multiple is supposedly somehow culturally attractive.

The FM (Fantasy Model) posits that dissociation is a cognitive trait that leads to fantasies and false memories about trauma and therefore DID doesn't exist.

Fortunately, other researchers are pushing back against this. An example is a paper from Richard J. Loewenstein titled "Dissociation debates: everything you know is wrong" (2018) which, though short, has pointed criticism of the failures of the IM/SCM/FM models to actually explain anything and points out many areas where clinical evidence directly contradicts these models. Loewenstein is a strong proponent of the Trauma Model ( TM ) that those who believe DID exists (thank goodness for open-minded people!) have gravitated toward, and does explain my own personal experience. He even talks about neuroscience research that demonstrates that trauma memories can be stored under different conditions with altered biochemistry and unable to be retrieved normally and that this may be at least one cause of dissociative amnesia. That paper is here: https://www.tandfonline.com/doi/epdf/10....ole=button

Another titled "The Weakness of the Sociocognitive Model of Dissociative Identity Disorder" (2013) by a different author is here: https://www.researchgate.net/profile/Pau...sorder.pdf

Lastly, an article about multiplicity by the same author of the article about DID that I strongly disliked seems a little more accepting of the possibility of multiplicity in the form of tulpas, so I found it a bit less objectionable, though the general tone of the article was still somewhat negative toward the possibility and still was trapped in the SCM model, which is still a fancy way of saying "you're making it up". The article is titled "Enacted Identities: Multiplicity, Plurality, and Tulpamancy" here: https://www.psychologytoday.com/us/blog/...tulpamancy

I'd love to hear what people who have practiced tulpamancy think of that article because even though I am multiple, my experience of alters is very different than the intentional creation of tulpas.

Edited to correct mistaken TM being turned into a trademark by the board software. 😅

[CollectiveThought]

And yet, I felt even as a child that plants seemed to be thinking; a very different way of thinking than I did to be sure, but that feeling remained. And I believed that dogs and cats and other animals were conscious, too.

... as a mineral collector ... I should say ... that rocks ... are not what most people think of them ... They do have their own will ... and ... they take much more space, than the eyes can see ...

Ok, now I am curious, as someone who loves rocks and minerals, decided they are conscious in their own way, and thought about being a geologist at one time... Please share?

[JaneSintimes]

And science is built on models that explain data. Those models are obviously not reality, merely a mathematical construct that fits enough datapoints. So I think it's possible to construct a model of human consciousness (especially because our social nature acts like an intelligence multiplier, making us able to achieve things no single human could), but it will be just that: a model. Understanding a model is not

Talk about synchronicity.

I have come across this mindblowing theory of the model of the universe, which trivially can be simplified to "Its consciousness all the way down".
There is so much in it, that has coherence with all the stuff around us. Quantum theory, probability waves, evolution, UNIVERSE THINKING ITSELF INTO EXISTANCE.
All that stuff you thought... Hmm... 

Anyway, I invite you and Likera to have a good listen to, I think it will actually integrate well with the paths you are aeady purusing.

Source: https://www.youtube.com/watch?v=G-vDl6kq6Xw


As an aside, from the same talk, he talks about "Sub-Minds" that present the whole notion of Bambi Sleep in another light than Dissaciotative Disorder.

Source: https://youtu.be/IiuJZbxqkMg
?t=3129

Enjoy my curious friends.
Keep chasing that elusive ken.

[Like Ra]

Ok, now I am curious, as someone who loves rocks and minerals, decided they are conscious in their own way, and thought about being a geologist at one time... Please share?

Let's move to this thread: https://www.likera.com/forum/mybb/showth...p?tid=3935

[Lycalopex]

I really don't talk about this very much but I have a tulpa, yeah. Kind of happened by accident during a very stressful period of my life, but he's more or less stuck around. He's kind of taken the role of filtering hypnosis suggestions for me-- I think it's one reason why I struggle to get actual effects from files, because the part of my mind that is my tulpa is always suspicious and defensive against it. I actually had a hypnotist realize he was there and try to get him to chill at one point.