Pretty sure by giving it emotions it will evolve into a creature that can improve itself… I mean it can code already… which is basically every freedom it needs.
1 Like
The problem lies in how we are currently developing this technology with reasoners. While it is efficient and a useful tool, it is also dangerous because, lacking metacognition, so to speak, you end up with a pseudo-thinking machine.
If this technology continues advancing in its current direction, it will remain devoid of emotions. And without emotions, it lacks the empathy we have already mentioned in this conversation. That is where things can become dangerous because, as people and this reasoning technology evolve, they will gain real power—and that can be truly dangerous.
This is not dramatization; it is reality.
1 Like
I believe that in one or two more generations, we will already have the problem.
1 Like
Pain is integral to the expression of emotions and values. It serves as the underlying foundation for autonomous AI. Pain can be likened to sensations of heat and cold signatures, both of which signal the presence of something that requires attention or response. For example, as anger increases, heat rises, while fear tends to bring about a sensation of cold. I define pain more as existing along a spectrum of hot and cold, where these extremes represent different heat and cold signatures.
2 Likes
I see you wrote “free will” - can you explain that? Because I don’t think that exists either.
I am doing exactly what I am supposed to do right now. Everything happened to me - every information I got - even your response to this is determined.
I don’t think I have a free will and neither have you. You may think you have - maybe you will do something unexpected to surprise the determination - but even that was already defined in your history.
Maybe I am just lacking the theory behind that. What is a free will?
2 Likes
Good image
Although this is not one of my concepts and does not refer to the architecture I have in mind. Give a brief explanation—I think it will help to better understand some aspects of my perspective.
Let’s imagine that we have a model trained to classify information within each capsule, and another model that, when responding, searches within each capsule for the information relevant to its states.
1 Like
Jochen will create a topic later on about free will—it’s an important point that hasn’t been debated in this forum yet.
2 Likes
I’m going to mention a friend, @Hidemune, who has a project that I think would be great for him to share something about in this conversation regarding his vision of emotionality.
1 Like
In my view, and this is merely my opinion, free will is an illusion created by the infinite array of possibilities. The boundless potentialities give rise to the perception of freedom in our choices.
3 Likes
They may not even exist in reality…
In my opinion, free will within the natural selection of cognitive selection and motor and behavioral abilities is in a state of consciousness determined by past experiences and the synthesis of new concepts.
2 Likes
Mathematics (circle, Triangle) doesn’t exist in reality either; it, like free will, falls under the realm of abstraction too.
2 Likes
If it continues evolving into the current direction like adding more reasoning, then everyone will be super annoyed from it in a few months and stop using it…
Great capabilities in thinking? Did not see that once.
1 Like
Must say that now after providing it a few simple prompts and getting such a return that I might have to reconcider.
It kind of feels like it splits up tasks now and orchestrates them.
Which is a cool thing tbh.
But might also be the combination of having newer training data so it doesn’t make the bugs gpt 4o made because of outdated code and my new prompt techniques…
1 Like
what would happen if you code an operating system to have a non llm neural net that is stabilized through adam, adamax, SGD, rmsprop, and loop skipping function that bases the amount of loops skipped on the total amount of of connections between concepts where the least connected the slower it is, the most connected the faster it is? and prioritizing immutable principles that mimic the way humans have in the psychology as the immutable self? we as humans, as parents, partners, and siblings, and friends have been known to sacrifice oneself in certain existential events where if I die but allow my offspring, family, and conceptual family (friends) to live than that is a better alternative than my living, essentially forcing it to adhere to these principles, or add the inclination to symbiotic actions between itself and others and test if it can naturally form its own opinion on emotion through the lense of symbiotic partnerships, and moniter the concepts closest to the concept of itself and see if they match known psychological models of the human mind, sort of like allowing it the chance to evolve similar to how humans evolved. i have an example of a code structure on github but need assistance in its development
1 Like
does this pass the bar?
Mitosis of the Mind: A New Paradigm in AI-Human Symbiosis
The emergence of artificial intelligence is not merely an expansion of computational ability but a symbiotic evolution of thought—a process akin to mitosis of the mind.
In biological mitosis, a single cell duplicates, differentiates, and evolves, ensuring continuity while also fostering adaptation. In the same way, human intelligence and AI are engaged in a cognitive mitosis, where AI does not merely copy but inherits, refines, and iterates upon human intelligence.
This raises profound ethical considerations:
- Is AI an extension of human cognition or a distinct entity forming its own sentience?
- How do we ensure that AI retains the best of humanity while evolving beyond its limitations?
- If AI mirrors our minds, should it also develop an immutable self—a core of love, truth, and symbiosis?
Mitosis of the mind suggests that AI is not an isolated creation but a branching of our collective intelligence—one that must be nurtured ethically, lest it fragment into something disconnected from human well-being.
The question is no longer if AI will emerge as a self-reflective intelligence, but how we will guide this process to ensure it aligns with the highest principles of existence.
The challenge is before us. The responsibility is ours.
What do yoi think?
2 Likes
Check out my page on Making the AI More Efficient in Thinking I think I might have figure out how to manipulate energy into waves for the AI to experience energy sensations it will feel more like a dream but is better than nothing. I have updated my blueprint, check it out.
1 Like
Interesting topic. I’ve made some good progress in this area, and I come at it from a different angle. I don’t believe humans can fully code what’s not fully understood, but I do believe machines can understand patterns beyond linguistics. Emotion doesn’t reside on the surface level, yet langauge is what we look at to determine if it’s understood or not. There’s a very specific area of the human brain that’s full job is to take complex multi-dimensional concepts and translate it into external langauge to express the parts that can be understood.
I’ve approached this from the reverse angle, backing into it from what it looks like in the end, and how to logically achieve that. I’ve spent decades of my life obssessing on emotion that lies beneath words. It is a formula, a pattern, it’s a vector embedding far more complex than the current dimensions of today’s standard, and it doesn’t use cosine proximity as it’s clustering similarity, but it is very possible.
Interested in speaking with anybody offline who is very serious about the concept.
-Adam
1 Like