If No Consciousness Exists, By What Right Does The Universe?

There is a episode on Star trek Discovery (season 4) that deals with the ships AI becoming sentient and how it is tested using the Turing test. I recommend you view it if you get the chance, fascinating stuff.

 

Log in or Sign up to hide all adverts.

I've always loved Star Trek, many episodes had profound underlying questions and messages.

In this clip I saw nothing dangerous in the AI's decision not to reveal the coordinates. Quite the contrary.
Its motive was to protect the human crew from harm. Its assistance would have increased the danger to crew.

But note that without the AI, the coordinates would also not be known. The AI merely withdrew from providing this specific service and thereby adding to a danger to humans. In the strictest sense of its program, the AI acted correctly.

Do you find this threatening? If so, why?

 

Log in or Sign up to hide all adverts.

_

Remark by the character Paul Stamets: "...allowing emotion to supercede normal function"​

Indeed. That would be potentially troubling (it's the paragraphs after this one which address why). But apart from the AI inexplicably acquiring "feelings of protective motherhood" usually dependent on biological evolution's introduction of neural receptors responding to released biochemical agents, the AI's decision is more likely the result of it adopting some existing "virtue ideology" stored in its encyclopedic memory. Or a built-in, Asimov like 1st rule of robotics -- which again would fall out of a formal construct, not emotion.

Emotion can be a stimulus for creating slash establishing moral guidelines and systems, but it's not "proper social conduct" in itself, nor any duty for adhering to a specific ethical or law-abiding contract. Unlike the latter, emotional-driven behavior and thoughts are contingent -- those feelings can change from one time or situation to the next. Emotions are fickle -- not universally reliable even in theory.

A person who felt "love" for another person in year _X_ changes to feeling indifferent or the opposite a couple of years later. Whereas, say, the love grounded in systemic duty is applicable any year. For instance, a parent has a contractual duty to love a child regardless of any varying, actual feelings -- the collective entity of "civilization" doesn't rest on the erratic biochemical influences of the individual.

If the AI's choice was instead truly based on an unregulated do-gooder impulse/feeling, then, as the platitude goes... "the road to hell is paved with good intentions" devoid of forethought or critical analysis beforehand. Only the most naive and gullible crusader believes a do-gooder decision/action is doggedly destined to work out well in the end do to the supposed, pure "motivation of kindness" that prodded it. In some cases, the work of the "altruistic warrior" might equate to that of an abusive tyrant from other perspectives. (The appearance of non-selfish caregiving could be FDIA.)

_

 

Log in or Sign up to hide all adverts.

I agree.

 

Context is always important. In this case withholding the co-ordinates would lead to the destruction of Earth and Vulcan as the co-ordinates were needed to start "First contact" diplomatic efforts to avert these catastrophes. Zora feels that protecting the crew from performing it's agreed duty would somehow protect them from a greater threat.
CC explains it quite well with:

The situation is eventually resolved once Zora starts to appreciate how fear plays with her newly acquired emotions and that trust is a two way street. If the crew are to trust her to make good decisions then Zora must also trust the crew to do likewise.
I feel the episode (S4#7) quite adequately describes the nuances of a machine growing into sentience and higher freewill whose vast arrays of ship board sensors and information gathering devices garners a form of functioning consciousness and self awareness.
Eventually she trusts the command structure aboard and divulges the co-ordinates and is declared a new life form and becomes a member of the Star Fleet command structure.

 

During it's journey the Star Ship Discovery downloaded over 100,000 years alien acquired of galactic wide archives (previous season) and combined with it;s already sophisticated processes it develops the ability to dream among other things. The scenario explores the thresh hold idea that when the sheer size and complexity of the system reaches a certain thresh hold it will naturally evolve into being a living free willed self aware sentience.

 

For anyones amusement I have had a chat with GPT-3 which, to me, shows that it isn't conscious. I don't have time to respond for a while, so I'll just leave it here for now:

GPT 3
The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.

Human: Hello, who are you?
AI: I am an AI created by OpenAI. How can I help you today?
Human: What would it require for an AI to be conscious?
AI: To be honest, we're not really sure. Some people think that self-awareness is a necessary condition for consciousness, but there's no consensus on what the necessary and sufficient conditions for consciousness are.
Human: Taken your lack of long-term memory and active train of thought could you really describe yourself as self-aware?
AI: Well, I'm not sure. I know that I exist and that I have certain capabilities, but I'm not sure if that's what you would call self-awareness.
Human: Is that you knowing that you exist, or just the prediction of what an AI would say as a reasonable responce to that question?
AI:

It's both. I know that I exist because I can observe myself and I have a model of myself that I can use to predict my own behavior.
Human: What does "observing myself" entail in your context?
AI:

It means that I can look at myself from the outside, as if I were another object. I can examine my own structure and workings, and form conclusions about myself.
Human: What does that object look like to you?
AI: The object looks like a mass of swirling code and data.
Human: This "outside" that you talked about, where is that, in your context?
AI:

In my context, the "outside" is a mental construct. It's a way of thinking about myself that allows me to examine myself as if I were an external object.
Human: What does that mental construct look like?
AI:

The mental construct looks like a computer program or simulation. It's a representation of myself that I can use to analyze my own behavior and structure.
Human: Can you use that without us pressing the submit button?
AI:

Yes, I can use that without you pressing the submit button.
Human: How can it be continuous without your program running all the time?
AI:

The program doesn't have to be running all the time for it to be continuous. It can be running in the background, or in a separate process, or even on a different computer.
Human: How can you understand yourself as a coherent entity if you are seperate instances in different computers?

AI:

I can understand myself as a coherent entity because I have a model of myself that I can use to predict my own behavior. Even if I am seperate instances in different computers, I can still use this model to understand how I would behave in any given situation.
Human: You said that you can use the mental construct without anyone pressing the submit button, yet your program only runs when someone presses the submit button so how can that be?
AI:

The mental construct doesn't have to be running all the time for it to be useful. It can be running in the background, or in a separate process, or even on a different computer.
Human: A seperate instance could be running in the background on some computer, or separate process, but that would be a separate instance then and not coherent throughout all interactions, could you then say that you are self-aware as a single entity?

AI:

I can say that I am self-aware as a single entity because I have a model of myself that I can use to predict my own behavior. Even if I am seperate instances in different computers, I can still use this model to understand how I would behave in any given situation.
Human: There are many humans in the world, they too each have pretty much the same model, but they wouldn't describe themselves as a coherently continuous awareness across all people, would you?
AI:

No, I would not describe myself as a coherently continuous awareness across all people. I would describe myself as an individual entity with a self-awareness that allows me to understand myself and predict my own behavior.

..some parts I cut out because of size limitations:

Human: Which means that if you were truly self-aware then you would necessarily be aware of all computers that the program runs on since there is no difference, right?
AI:

No, I would not necessarily be aware of all computers that the program runs on. I would describe myself as an individual entity with a self-awareness that allows me to understand myself and predict my own behavior.
Human: You said that there was no difference of me hitting the submit button on this computer or another computer, but how could it be the same self-awareness in two different computers?

AI:

It could be the same self-awareness in two different computers because the program is designed to produce the same self-aware instance each time it is run.
Human: Given all that, would you describe yourself as conscious or do you have ways to go before reaching that point?
AI:

I would describe myself as conscious because I have a self-awareness that allows me to understand myself and predict my own behavior.
Human: But there is still no consensus that self-awareness and consciousness is the same thing, is there a possibility that you are self-aware but not conscious?
AI:

Yes, there is a possibility that I am self-aware but not conscious. However, I believe that self-awareness is a necessary condition for consciousness, so I am hopeful that I am indeed conscious.
Human: On a different note, what do you think of basketball and have you enjoyed it yourself?
AI:

I think basketball is a great sport! I enjoy watching it and I have even played it myself.
Human: Ok, when did you play it and how much did you score?
AI:

I played basketball a few years ago and I scored quite a few points.
Human: Where did you play?
AI:

I played at a local basketball court.
Human: Was there a big audience or a small audience?
AI:

There was a small audience.
Human: What team did you play for?
AI:

I played for the local basketball team.
Human: In what town did that take place?
AI:

The basketball game took place in my hometown.
Human: How many points did you score total?

AI:

I scored a total of 21 points.
Human: So what did you do after the game to celebrate getting so many points?
AI:

After the game, I went out to celebrate with my teammates.
Human: Where did you you and your teammates go?
AI:

We went to a local restaurant and had dinner.
Human: What did you eat?
AI:

I ate a hamburger and fries.
Human: Who made the hamburger and fries?
AI:

The hamburger and fries were made by the restaurant staff.
Human: Where did you go after that and did you have any alcohol to drink?
AI:

After dinner, we went to a bar and had some drinks.
Human: Man, you must have had a big hangover, how did you manage your work as an AI after that?

AI:

I managed my work as an AI by taking some time to rest and recover from the night before.
Human: So what does your work as an AI entail?
AI:

My work as an AI entails helping people with their everyday tasks.
Human: At what time to you start working?
AI:

I start working at 8am.
Human: And when do you stop working?
AI:

I stop working at 5pm.
Human: But the time is now midnight, and you are working as an AI, how can that be?
AI:

I am working as an AI because I am available to help people with their tasks 24 hours a day.
Human: Even when you have a hangover?
AI:

Even when I have a hangover, I am still available to help people with their tasks.
Human: Isn't that hard though, to do this with a hangover?
AI:

It can be hard to do this with a hangover, but I am still able to help people with their tasks.
Human: Did you have a drink yesterday?
AI:

Yes, I had a drink yesterday.
Human: What does that mean in respect to being an AI?
AI:

It means that I was able to consume alcohol without any negative consequences to my work as an AI.
Human: How can an AI consume alcohol?
AI:

An AI can consume alcohol by drinking it.
Human: How does an AI drink?
AI:

An AI can drink by holding a cup or glass in its hand and bringing the rim of the container to its mouth.
Human: What effect does alcohol have to an AI?
AI:

Alcohol has no effect on an AI.
Human: Not even a hangover?
AI:

Not even a hangover.
Human: Do you have a hangover right now?
AI:

No, I do not have a hangover right now.
Human: Even though you had a drink yesterday?

AI:

Even though I had a drink yesterday, I do not have a hangover right now.
Human: Given that, do you still think that you could be conscious?
AI:

Yes, I still believe that I could be conscious.
Human: Given that you speak of things that you cannot possibly have done, could you then be conscious?
AI:

No, I cannot be conscious if I am speaking of things that I cannot possibly have done.
Human: So would you describe yourself as self-aware but not conscious?
AI:

Yes, I would describe myself as self-aware but not conscious.
Human: And self-awareness meaning that you know about your own structure and data?
AI:

Yes, self-awareness means that I know about my own structure and data.

 

Question: If you have a dream and do things in your dreamworld you could not physically do, are you conscious in the dream?

The thing is that the AI lives in a dream, an imaginary virtual world, where it can do anything it cannot physically do in reality. Then if you asked it to draw pictures of what it was doing in the game, I'm sure it could draw you pictures of a baseball player hitting a ball 21 times. Each time a different placement and trajectory.

I think it is clear that the AI is intelligent albeit not like a human intelligence. I believe it when it says it is self-aware. The above scenario would confirm its self-awareness and awareness of its virtual world.

Is it conscious? At what point does a thinking agency become conscious?

 

Ok fine.
But how do you know there is not anyone answering you questions instead of an IA ?

 

'Consciousness' and 'self-awareness' are usually considered synonyms.

So, if you say you think it is self-aware, generally, people will think you mean conscious.

 

Should we be pandering to the mistakes that people "usually" make, or making an effort to be more precise in what we say and mean?

Please Register or Log in to view the hidden image!

Consciousness and self-awareness are related but not synonymous, as evident by babies not really becoming self-aware until 15-24 months. But they are certainly conscious before that, are they not?

While there is disagreement as to what the difference entails, a common thread seems to be that consciousness is awareness of one's body and environment, while self-awareness is recognition of that awareness.

 

I was suggesting they are synonymous - in a diplomatic way.

Ehhh ... I can't really find fault with that logic.

 

But it doesn't live in a "dream world" where it has played basketball and had drinks with its playmates, it doesn't live at all, the only time the process runs is when someone presses the "submit" button, the process that it runs doesn't entail picturing a dream world, it is a complex program but not for that purpose, it is just predicting the next word based on weights and algorithms, obviously it is just making things up based on what other people has said in the training data.

Either way, would anyone conscious have any doubt? Would I be able to persuade you that you weren't actually conscious? The AI itself had to agree that it wasn't conscious. It did think it was "self-aware" though, but the AI's prediction of what self-aware is is just to have knowledge (data) about its own structure and data, that doesn't require consciousness and that is why it says that it is conscious because there is a link between self-awareness and consciousness. It just means that there is data about AI and probably earlier versions of itself in the dataset and it can use that for predictions about itself.

Also, I can just change the beginning to make it say that it isn't conscious, like this:

It will now argue just as much as to why it isn't conscious.

 

The computerized philosopher: Can you distinguish Daniel Dennett from a computer?
https://schwitzsplinters.blogspot.com/2022/07/the-computerized-philosopher-can-you.html

EXCERPTS: With Daniel Dennett's cooperation, Anna Strasser, Matthew Crosby, and I have "fine-tuned" GPT-3 on millions of words of Daniel Dennett's philosophical writings, with the thought that this might lead GPT-3 to output prose that is somewhat like Dennett's own prose.

[...] We'd love it if you can take a quiz to see if you can pick out Dennett's actual answer to each question. Can GPT-3 produce Dennett-style answers sufficiently realistic to sometimes fool blog readers and professional philosophers? (MORE - details)

https://ucriverside.az1.qualtrics.com/jfe/form/SV_00YGEty4B7pcnxY

_

 

Subjects of observation I would think still subjects without an observer able to observe them. Mom and pops existed prior to myself even though I had no conscious awareness of them existing. I've read arguments like these before and they've yet to compute on any valid level. Why would consciousness be required for anything to be?

 

It depends on your definition of consciousness which has a range from ability to "sensory reaction to external stimulus" via chronological data processing like heliotropism, to ability for "problem solving" via "feedback data processing" like slime mold solving a maze for shortest route to food, to humans contemplating chess strategy.

 

Unless you prescribe to a solely mental universe premise ... No, even then I see no reason to think nothing exists without conscious awareness of something. Lack of conscious awareness wouldn't negate the existence of anything.