3 ms·

How is it any more naive than thinking consciousness can arise on it's own from a box of carbon switches (aka a brain)?

chromanoid 5y ago
> from a box of carbon switches "To find out, David Beniaguev, Idan Segev and Michael London, all at the Hebrew University of Jerusalem, trained an artificial deep neural network to mimic the computations of a simulated biological neuron. They showed that a deep neural network requires between five and eight layers of interconnected “neurons” to represent the complexity of one single biological neuron." https://www.quantamagazine.org/how-computationally-complex-is-a-single-neuron-20210902/ https://www.quantamagazine.org/how-computationally-complex-i... Beside this there is also https://en.wikipedia.org/wiki/Ephaptic_coupling https://en.wikipedia.org/wiki/Ephaptic_coupling, chemical reactions and maybe even quantum effects. And then you have to think about https://en.wikipedia.org/wiki/Panpsychism https://en.wikipedia.org/wiki/Panpsychism and so on.
- naasking 5y ago
  > They showed that a deep neural network requires between five and eight layers of interconnected “neurons” to represent the complexity of one single biological neuron. So basically, a neural network "neuron" is about an order of magnitude less complex than a biological neuron. Furthermore, GPT-3 has at least 3 orders of magnitude fewer nodes than a human brain has neurons, and multiplied by the order of magnitude difference between that leaves 4 orders of magnitude in complexity between best in class neural networks and human brains. Will bridging this gap reproduce all human complexity, including consciousness? Hard to say. Human neurology has a lot of specialization too, and general systems typically need a lot more resources to emulate specialized systems (think ASICs or GPUs vs CPUs). So there may be further orders of magnitude needed to close that gap fully, but given how scary competent GPT-3 is, you can see how GPT-N that has 4 orders of magnitude more complexity might be scary competent, bordering on "intelligent".
  - chromanoid 5y ago
    You are aware that we are not even able to simulate C. elegans with its ~1000 cells including "only" 302 neurons?
jqpabc123 5y ago
It's not naive to acknowledge existing examples. But it is naive to assume that nothing more than "carbon switches" give rise to something we can't even begin to fully define yet.

> from a box of carbon switches "To find out, David Beniaguev, Idan Segev and Michael London, all at the Hebrew University of Jerusalem, trained an artificial deep neural network to mimic the computations of a simulated biological neuron. They showed that a deep neural network requires between five and eight layers of interconnected “neurons” to represent the complexity of one single biological neuron." https://www.quantamagazine.org/how-computationally-complex-is-a-single-neuron-20210902/ https://www.quantamagazine.org/how-computationally-complex-i... Beside this there is also https://en.wikipedia.org/wiki/Ephaptic_coupling https://en.wikipedia.org/wiki/Ephaptic_coupling, chemical reactions and maybe even quantum effects. And then you have to think about https://en.wikipedia.org/wiki/Panpsychism https://en.wikipedia.org/wiki/Panpsychism and so on.

naasking 5y ago
> They showed that a deep neural network requires between five and eight layers of interconnected “neurons” to represent the complexity of one single biological neuron. So basically, a neural network "neuron" is about an order of magnitude less complex than a biological neuron. Furthermore, GPT-3 has at least 3 orders of magnitude fewer nodes than a human brain has neurons, and multiplied by the order of magnitude difference between that leaves 4 orders of magnitude in complexity between best in class neural networks and human brains. Will bridging this gap reproduce all human complexity, including consciousness? Hard to say. Human neurology has a lot of specialization too, and general systems typically need a lot more resources to emulate specialized systems (think ASICs or GPUs vs CPUs). So there may be further orders of magnitude needed to close that gap fully, but given how scary competent GPT-3 is, you can see how GPT-N that has 4 orders of magnitude more complexity might be scary competent, bordering on "intelligent".
- chromanoid 5y ago
  You are aware that we are not even able to simulate C. elegans with its ~1000 cells including "only" 302 neurons?

jqpabc123 5y ago

It's not naive to acknowledge existing examples. But it is naive to assume that nothing more than "carbon switches" give rise to something we can't even begin to fully define yet.