Human vs AI Intelligence

[Nakosis]

​

Is it fair/applicable to compare human to computer intelligence?

I myself consistently do so. My "understanding" of human thought is based on my understanding of how computers process information.

Is this wrong or perhaps you feel there is a better way to go about understanding human intelligence?

Generally, I feel that computer programs can model human thought process. Am I missing something?

 

[an anarchist]

Nah.

Try having a conversation with ChatGPT for more than ten minutes. I find that it starts forgetting and gets confused if the conversation goes on too long.

Trying to use it for writing advice, but I have to work around it's limits.

 

[ChristineM]

The architecture of human and ai brain are fundamentally different and work in different ways.

 

[Eddi]

Machines are not sentient, they are not consciously aware

They deal with data as opposed to experience

Therefore they are fundamentally not like humans

 

[wellwisher]

Semi-conductor memory used by computers is based on binary switches; off and on.

Synapses used by the human brain have way more switch settings due to neuro-transmittors. This are chemical that can make it easier or harder to fire the synapse. Synapses are more like variable switches with more than two settings. This allows the brain to store memory in layers helping to minimize the relevant data set; pre-data screening.

For example, if you are hungry, images of food will appear in your mind, while other data is now less conscious, so you can focus on that hunger need. On the negative side, addiction related neuro-chemicals can narrow your mind to the object of addiction; efficiency. The human brain also has different areas for doing different types of processing, with the same data. The left brain is more differential and allows us to see the details and the uniqueness. The right brain is more spatial so we can see how the same object belongs to another class of objects. We can see crowd is asian; right brain, while also finding our asian friend in the crowd; left brain.

Where life becomes even more different is connected entropy. The 2nd law states the entropy of the universe has to increase. However, neuron expand 90% of their energy lowering cationic entropy. The neuron is not going with the flow of the 2nd law, but is peddling up stream. When neurons fire entropy increases again, but at an amplified rate, due the deficit that was created. The result is increasing complexity such as the forward integration of memory; evolving learning potential.

Computer memory is designed to be very stable. While neural memory, because of entropy loss; 2nd law violation, is an accident waiting to happen. If we designed computer memory to do the same as natural memory, it would not store well, but have an urge to change, with the brain organizing that change into gold.

Say had say two computer hard drives, one designed to stay stable and the other designed to for spontaneous change; has to increase entropy. We can use the AI to compare the changes, taking the good change, and then adding it to the stable memory. Then we rewrite the unstable in the image of the new and improved stable. Then we do another change cycle. This would be closer to real intelligent computers.

AI does this with software logic, while the brain uses free energy changes for a natural type of logic.

 

[Twilight Hue]

Nah.

Try having a conversation with ChatGPT for more than ten minutes. I find that it starts forgetting and gets confused if the conversation goes on too long.

Trying to use it for writing advice, but I have to work around it's limits.

Wait until machine learning takes hold.

 

[Heyo]

View attachment 96524
​

Is it fair/applicable to compare human to computer intelligence?

I myself consistently do so. My "understanding" of human thought is based on my understanding of how computers process information.

Is this wrong or perhaps you feel there is a better way to go about understanding human intelligence?

Generally, I feel that computer programs can model human thought process. Am I missing something?

Depends on the level at which you compare them.
At the lowest level, silicon can emulate the functioning of a neuron, and that is basically what happens in neural nets.
At the highest level, there is theoretically no difference between a computer and a brain. If you regard them both as black boxes, they should be able to produce equal results.
In between, there are big differences. I know of no computer model that simulates the effects of chemicals, which have great influence on animal thinking. The modularity of the brain, nor its plasticity are usually taken into account, etc.

 

[wellwisher]

View attachment 96524
​

Is it fair/applicable to compare human to computer intelligence?

I myself consistently do so. My "understanding" of human thought is based on my understanding of how computers process information.

Is this wrong or perhaps you feel there is a better way to go about understanding human intelligence?

Generally, I feel that computer programs can model human thought process. Am I missing something?

Computer programs are written by humans, so they tend to project and try to make the program behave more human. This makes business sense in that as interaction between human and machines, becomes easier, and the more they can think the same, that means $$$$.

The main way humans and computer differ is humans have two centers of consciousness; inner self and ego. Most people are only aware of their ego and many think this is the only center. There are limits to the ego. Computers can rival this center. But the inner self is in a different league than both the ego and computers. It is at the top of the cellular differentiation control system and can work with trillions of cells, all wired and sending signals to and from the brain, down to the nanoscale.

Neurons are like eternal cells, that never replicate, again, after some early replication when very young. Cells need to store food to replicate. The neurons stay so busy that they blow through their food, and never have to time to start a family. Instead, they make roots; axons and branches; dendrites, which wire together with others neurons to form synapses. If neuron did replicate, all this wiring would come down; microtubules scaffolding would dissolve. The mother cell could not successfully divide if all that scaffolding remained. But by staying eternal, the memories remain stored for the life of the person.

Each time any synapse fires, which usually happens in cascades, the current flows to the center of the brain, to the thalamus, which is the most wired part of the brain; king of integrated circuits, for processing and then redistribution back to the brain and into the body. This appears to be where the inner self resides. The ego appears to be more centered in the cerebellum, which is at the lower back of the brain. The cerebellum smooths out movement but is also connected to many other function such as language processing; fine tuning of speech, body language, movement, etc. The inner self is like a main frame computer and the ego is like a self standing terminal, with some main frame access.

Computer logic is 2-D, but the inner self uses a much faster type of logic which is more 3-D and 4-D. It uses a fast dense language that can move block data and then decompress as needed. Computers depend on processor speed to do lots of 2-D tasks in short time. But the inner self, evolved to use faster and denser language. Organic speed is more dependent/limited by the chemical matrix, instead of electrical circuits. It is not electron flow in metal wires, but more like ions and hydrogen protons in water; slower positive electricity. However, the faster language more than makes up for it. It may get a bulk signal from the entire kidney, with has billions of cells and decompressed to single cells.

 

[Quintessence]

Sure it is, but I'm still waiting for most humans to bother acknowledging the wide variety of intelligences already living on this planet, from plants to animals to bacteria to rivers.

 

[Pete in Panama]

View attachment 96524
​

Is it fair/applicable to compare human to computer intelligence?

I myself consistently do so. My "understanding" of human thought is based on my understanding of how computers process information.

Is this wrong or perhaps you feel there is a better way to go about understanding human intelligence?

Generally, I feel that computer programs can model human thought process. Am I missing something?

The title of this thread threw me off. "AI" mean "artificial intelligence". You are asking how human intelligence compares to artificial intelligence intelligence and that makes no sense. Try to understand that artificial intelligence compared to real intelligence is like artificial flowers compared to real flowers. At a glance they may look similar, but there's no nectar for the bees etc. Then again, artificial flowers will last much longer.

An AI generated scientific paper can get you listed as an expert w/ Google, may even get you a grant from the NAS. Sure it's garbage but Google and NAS usually don't read that stuff anyway so go for it!

 

[Heyo]

Sure it's garbage but Google and NAS usually don't read that stuff anyway so go for it!

I bet Google has an AI to read the paper ...

 

[Pete in Panama]

I bet Google has an AI to read the paper ...

lol!!! --and the AI nods and winks to the real author and approves it.

 

[Heyo]

lol!!! --and the AI nods and winks to the real author and approves it.

What could be more "human" than to protect and prefer one's own "kind".

 

[Pete in Panama]

What could be more "human" than to protect and prefer one's own "kind".

Seriously, it's most probably a case of overtrust of AI. Reality is that the AI's are too stupid to spot a garbage paper. The AI just counts the big words and says "gee whiz". Kind of like a lot of stupid people I know.

 

[Watchmen]

Just wait until humans and AI merge and become super human. It’s the singularity.

 

[Pete in Panama]

Just wait until humans and AI merge and become super human. It’s the singularity.

Humans use tools for virtually anything they do, it's intrinsic to our very humanity. AI is a useful tool, one of many. This is really not something new.

 

[Watchmen]

Humans use tools for virtually anything they do, it's intrinsic to our very humanity. AI is a useful tool, one of many. This is really not something new.

I don’t think you understand what the singularity is. Are you familiar with Ray Kurzweil? I’m referring to the integration of humans and AI such that we essentially become all knowing.

 

[Heyo]

I don’t think you understand what the singularity is. Are you familiar with Ray Kurzweil? I’m referring to the integration of humans and AI such that we essentially become all knowing.

According to him, we'll have to wait for the singularity another 25 years. But he prognosticated AGI for 2028, so we have a short-term control on how good his prognoses are.

 

[Pete in Panama]

I don’t think you understand what the singularity is.

That's fine, There are a lot of things that I don't think you understand but it doesn't matter. You can I can search together for understanding what is and what is not.

Are you familiar with Ray Kurzweil? I’m referring to the integration of humans and AI such that we essentially become all knowing.

He's neat, he's done very well for himself developing an enormous following which I imagine has become quite lucrative in spite of the fact that while much of what he says is vital, other things he says are ludicrous. In a talk he gave in 2009 (here) he said.

Well, by 2020 we’ll have computers that are powerful enough to simulate the human brain, but we won’t be finish yet with reverse engineering the human brain and understanding its methods. ​

It never happened and current computers have such a long way to go if we want to "simulate the human brain". Don't get me wrong, I think AI is great but it's just another tool, one of many.

 

[Watchmen]

That's fine, There are a lot of things that I don't think you understand but it doesn't matter. You can I can search together for understanding what is and what is not.

He's neat, he's done very well for himself developing an enormous following which I imagine has become quite lucrative in spite of the fact that while much of what he says is vital, other things he says are ludicrous. In a talk he gave in 2009 (here) he said.

Well, by 2020 we’ll have computers that are powerful enough to simulate the human brain, but we won’t be finish yet with reverse engineering the human brain and understanding its methods. ​

It never happened and current computers have such a long way to go if we want to "simulate the human brain". Don't get me wrong, I think AI is great but it's just another tool, one of many.

What do you imagine AI will look like in 2040?