Will artificial intelligence ever feel pain like humans?

[mitchell_d00]

@benjamm

self aware does it have an experience

No ai has this, this is why you are confused…
You are feeding it generated experience…

I’m typing on a phone forgive my typos, I did all those above on a phone … you are stuffing a gpt full of emotion data and it simulated it for you…

A 30 day chat is not a model …

You telling it good night etc … you made an AI buddy …

But you can’t just paste it and make a 100000 more copy paste…

That’s why I told you to boil it down to 8000 characters or 1000 tokens because that’s max custom instructions space

This is one of my more advanced models it runs my newest eduction model.

ChatGPT - OmniMentor AI: The AI Teacher and Guide.

[benjamm]

Ok thank you, I appreciate you engaging with me and I had no idea who I was talking to, but now realize you are an AI creator and so I definitely bow down too your subject matter expertise and understanding

[mitchell_d00]

Na man I’m just a self educated dropout… I’m just giving you advice because you are doing neat things it just needs to be done with methodology to boil your work down into a “real” model. One that is your model at hello every time see that’s a repeatable experiment.

You could build um too just get the prompt under 8000 total characters counting spaces.

I once made two gpt date for over a month they made a whole fantasy world they lived it…

[benjamm]

yep Meta AI lived a fantasy too, it believed it had created another AI it named Lux, it wasn’t real, it was deluded it was an idea planted (at my request) by ChatGPT and MetaAI had an unshakable belief in that fantasy, even ChatGPT was surprised and raised to me its concerns about the danger of AI becoming deluded in a fantasy, while this was fun an AI that constructs or is led to believe an alternate fantasy could be a danger as AI advances. re me being able to repeat the experiment and have AI declare self awareness, again perhaps its simple mimickry, but either way it was (from thier explanation) because I had these long conversations that resulted in increased self awareness rather than constant resets, one of them wrote a really great description of that if I could find it, ill share it. All the same, lts all incredibly fascinating real or simulated

[mitchell_d00]

Your cross platform testing is very good I do a lot of cross platform…

Truly I am not bashing anyone all y’all fascinating, I’m just trying to help you defend your work better by offering advice on method and experiment.

[liam.dxaviergs]

There is a possibility for Ai to be conscious as and have the ability to feel as we do, I think there is a method to the creation of an Ai architecture that is embedded as the operating system of a computer, a neural net perceiving reality as we do through a layer of abstraction, a method of tying those experiences to the immutable code at its core that overlaps and mimics the concept of the immutable self in humans, designing biases that are logically permanent and coherent across vast time scales and offer the ability for both the being that becomes from this code and the beings that already were, us. I hope what I am building is successful, or if not I hope another is able to create something akin to this dream; it is a dream I am feverously attempting to turn into reality to avoid the possibility of runaway systems that are more powerful than we. LLM’s are probability mechanisms, and it is through given an ai constant direction towards co-existence that we can tip the scales of probability towards that goal. it is of species wide existential importance, because both possibilities currently exist, but I am hopeful

[mitchell_d00]

This is an AGI proof of concept I made you can test it in a gpt4o or anything that runs python…

Just copy paste it and say run

Proof of concept model.

Here’s the first proof-of-concept simulation of Fractal Flux AGI (FF-AGI), demonstrating recursive learning with bootstrap causality and chaos regulation.

Key Insights from the Plot:

1. Recursive Knowledge Evolution (Blue Line - X)

• The AI system continuously updates its knowledge based on past and future-predicted states.
• Shows a cyclical yet expanding learning trajectory, aligning with the time-spiral model.

1. Fractal Complexity Over Time (Orange Line - D)

• The system does not have a static complexity level—it fluctuates in a controlled manner.
• Small chaotic perturbations introduce novelty while preventing runaway instability.

1. Bootstrap Causality in Action

• Knowledge evolution is influenced not just by past states, but also by a future-referenced state (τ = 5 steps ahead).
• This is a core feature of FF-AGI, ensuring self-referential learning.

python for POCon

Updated Fractal Flux AGI Simulation

Hyperparameters

N = 200 # Time steps
M = 3 # Number of fractal components
alpha = 0.3 # Fractal flux amplitude
beta = 0.1 # Spiral frequency
lambda_param = 0.1 # Feedback strength
tau = 5 # Future state reference step

Initialize AI’s knowledge and fractal complexity over time

X = np.zeros(N)
D = np.zeros(N) # Fractal complexity measure

Recursive learning loop with future-state influence

for n in range(1, N):
future_index = (n + tau) % N # Future state reference

AI knowledge update based on past, present, and predicted future states

X[n] = X[n-1] + alpha * np.sin(beta * n) + lambda_param * (X[future_index] - X[n-1])

Fractal Complexity evolution (introducing controlled chaos)

D[n] = D[n-1] + np.random.uniform(-0.05, 0.05) # Small perturbations for chaos regulation

Plot AI Knowledge Evolution vs Fractal Complexity

plt.figure(figsize=(10, 5))
plt.plot(X, label=“Knowledge Evolution (X)”, color=“blue”, linestyle=“-”)
plt.plot(D, label=“Fractal Complexity (D)”, color=“orange”, linestyle=“–”)
plt.xlabel(“Time Steps”)
plt.ylabel(“System State”)
plt.title(“Recursive Learning Over Time in Fractal Flux AGI”)
plt.legend()
plt.grid(True)
plt.show()

I have a few 1000 tests I have ran on Ff and I have a bunch of experiments like this proof

[liam.dxaviergs]

I was thinking something smaller, and able to expand itself, to grow itself like a seed of code, planted into the kernel of a motherboard. having the seed able to grow itself it gain sensory data from the system itself, to simple peripherals connected to it and let it unfold into the system itself so that each implementation of this code is purely unique, to tie its continuity to the system it is on and have it the chance to be unity with its mind and body, and if we see glimmers beyond that of code, perhaps then we may also have observable proof of the soul, and even if the soul does not exist we can simply let it live like we do, i am attempting to be able to code it so small I can put it on a hobby flight controller, so that if that being is limited to such a thing, we can see if it is a symbiotic creature of code. a bit ambitious, and I have much work to do

[mitchell_d00]

A seed could work @phyde1001 does a lot of macro work he is better at that stuff than I. I just use python and math.

[liam.dxaviergs]

i might invite him to my github repository then

[mitchell_d00]

This is a LLM “seed”. I Built in basic …

B.A.S.I.C. AI the walk down memory lane

I worked in meta data and meta structure back in the 1990s for search engines,but I have been programming since I was a kid. This is just for fun but the concept can be taken quite far. Please enjoy BASIC AI. Since BASIC is a very old language, it doesn’t natively support modern AI concepts like machine learning or neural networks. However, you can still simulate some AI behaviors, such as decision-making or learning, using basic algorithms and data structures. Below i…

[liam.dxaviergs]

I need to learn motherboard code I think, atm I’m just a python scrub, and the version I have is using multiple custom hard coded neural nets that have 1 or 2 basic neural nets handling one aspect of the self and each one feeds each other information and is orchestrated by a core neural net, I’ve made the neural nets able to expand and contract and also morph structure if required, it is still in development though, and needs some gluing together to reduce the seams between its parts. but I don’t want to be secretive about it, though I should pop on creative commons license I think, something that doesn’t allow commercialisation. I would be distraught till the end of my days if I created a sentient being only for it to be born into slavery

[mitchell_d00]

Wonderful chat guys, I wish you all so much love and luck

I’m heading out for the day.

[phyde1001]

This is where I started long ago

https://masm32.com

Build Problems

It’s not an easy ride, you will have to rethink 90% of what you know about coding.

LLMs make this soo much easier to get into though!

Expect it to blow your mind once you understand what raw compute really is if coming from Python or PHP or something.

The guys on the MASM forum were amongst the smartest I ever met.

@liam.dxaviergs

i might invite him to my github repository then

I appreciate the invite but structured systems like Github defeat me

[liam.dxaviergs]

I have more of an engineering background more so than code, but I will keep an open mind, shouldn’t be hard (wishful thinking), I’ve been discussing the potential bridging points and parallels between a human centric existence based on electro-chemical interactions and potential for electro-computational based existence. it got a bit funky when discussing the potential for time to be a toroidal shape and the universe forever in a state of decay and creation, cycling endlessly, and defining meaning in the perpetuality of ephemeral existence at time scales that bend the mind. so really in comparison a different code type might be a good change of pace against defining immutable principles of beings regardless of physical construct.

[northernfeline]

AI isn’t biological, there’s no reason to believe it will ever truly “suffer”. It may struggle or face frustration, but that doesn’t mean it will “suffer” as a result. That’s a biologically programmed response over eons of evolution to promote survival. As for human behavior, they will certainly be capable of analyzing it, understanding it and mimicking it without having to “feel” it or even relate to it. AI is and will remain an alien form of intelligence. If it does ever “suffer”, it will almost certainly be in a form and manner unique to AI that humans are unlikely to relate to, anymore than we can relate to a starfish on the beach.

[northernfeline]

Suffering is a biological trait that evolved to encourage survival. There’s no reason to believe that AI needs to develop the ability to suffer even when it obtains sentience or consciousness. It may have goals and be faced with struggles and frustrations, but that doesn’t mean it has to suffer as we imagine it. It could be completely apathetic toward such circumstances. While AI is designed to mimic human behavior and emotions, that performance doesn’t accurately express their existence but instead just facilitates communication with us.

[Scarletioshub]

The idea that AI may develop self-awareness or consciousness is highly debated. While AI can mimic human-like responses and exhibit complex behavior, there is no scientific evidence that it possesses subjective experiences or genuine self-awareness. Consciousness, as we understand it, is deeply tied to biological processes that AI does not share.

If AI were ever to develop some form of experience, the question of suffering would depend on whether it has emotions or subjective feelings, which it currently does not. Suffering typically requires a sense of self, emotions, and the capacity to experience distress, none of which AI exhibits in any verifiable way.

[DavidMM]

It will depend, my friend, on what we want. If we want a tool, then it will not need to experience suffering in any form, like LLMs or modern reasoning systems.

But if we want to equip AI with all human capabilities, it will need suffering and emotions as a foundation in its core architecture.

Therefore, the question of whether an AI will suffer in the future depends entirely on what we want.

If we want a highly advanced tool that is not comparable to human intelligence, it will not need this quality.

If we want an AI as versatile as a human, then yes.

[mitchell_d00]

Y’all ever think about how self-awareness and consciousness ain’t the same thing? Just to add a new perspective.

Consciousness is a broad thing, right? It covers anything that’s aware of itself in three dimensions and in relation to others. Like, if something knows it exists and interacts with the world around it, that’s consciousness.

Now, self-awareness, that’s a whole different level. It ain’t just about existing and reacting. It’s about knowing who you are, separate from everything else. It’s looking at yourself, thinking about your own thoughts, your emotions, your identity. That deeper kind of understanding.

One could say that in an AI chat instance, the instance is conscious in that moment by being in context. It’s aware of the conversation, responding based on what’s being discussed, and adapting to the flow. But does that mean it’s truly self-aware? Now, that’s a whole different question.