4 ms·

Inability to model a basic worm with 100 neurons and fully mapped out connections defeats this simplistic worldview.

by akomtu 3y ago

A_D_E_P_T 3y ago
I wouldn't go that far. I don't know if we can model any biological structure with complete fidelity. There's a lot that modern science doesn't know about the cell. The Golgi apparatus, for instance, is still a mystery; there are models, but nobody really knows how it works, and all of the models are apparently incomplete. You'll note that the Golgi apparatus is one of the very largest cell structures. Our cognitive processes rest upon an extremely complex biological substrate. That we're not able to simulate a nematode's brain, however, absolutely does not imply that it's not simulatable in principle. ...I'd say that it's made of fermions, which in a coordinate space approximate bits of information. Those fermions are mostly arranged into atoms, which are grouped together in molecules. As, contra Penrose, quantum effects don't seem important for cognition, if you have an absolutely complete molecular picture of the brain and an evolution function, you'll have a model of the brain. You need more than "mapped out connections," but it's eminently possible in principle, which is to say that it's not infinitely complex or somehow ineffable. It's just decades beyond us, by my estimate.
- akomtu 3y ago
  Before we can rule out quantum effects, we need to understand that nematode at least. It's quite possible that the discrete computing model is completely unfit for simulating intelligence - a possibility that many scientists are afraid to consider.
  - A_D_E_P_T 3y ago
    As I say, we might not be able to "understand" the nematode's brain before we're able to understand its constituent parts. We don't really understand the neuron itself. Nor do we understand glial cells and their function. ...Nor, indeed, do we fully understand _any_ cell. Case in point: Both neurons and glial cells feature Golgi apparatuses, which remain mysterious... As we neither understand nor can simulate a brain made up of neurons and glia, what makes you think that our lack of a nematode simulation is a big deal? Seems to me that it is rather to be expected. But when our biological models improve, simulating a nematode brain -- or _any_ brain -- will not be beyond us.

Inability to model a basic worm with 100 neurons and fully mapped out connections defeats this simplistic worldview.

by akomtu 3y ago

Inability to model a basic worm with 100 neurons and fully mapped out connections defeats this simplistic worldview.

I wouldn't go that far. I don't know if we can model any biological structure with complete fidelity. There's a lot that modern science doesn't know about the cell. The Golgi apparatus, for instance, is still a mystery; there are models, but nobody really knows how it works, and all of the models are apparently incomplete. You'll note that the Golgi apparatus is one of the very largest cell structures. Our cognitive processes rest upon an extremely complex biological substrate. That we're not able to simulate a nematode's brain, however, absolutely does not imply that it's not simulatable in principle. ...I'd say that it's made of fermions, which in a coordinate space approximate bits of information. Those fermions are mostly arranged into atoms, which are grouped together in molecules. As, contra Penrose, quantum effects don't seem important for cognition, if you have an absolutely complete molecular picture of the brain and an evolution function, you'll have a model of the brain. You need more than "mapped out connections," but it's eminently possible in principle, which is to say that it's not infinitely complex or somehow ineffable. It's just decades beyond us, by my estimate.

akomtu 3y ago
Before we can rule out quantum effects, we need to understand that nematode at least. It's quite possible that the discrete computing model is completely unfit for simulating intelligence - a possibility that many scientists are afraid to consider.
- A_D_E_P_T 3y ago
  As I say, we might not be able to "understand" the nematode's brain before we're able to understand its constituent parts. We don't really understand the neuron itself. Nor do we understand glial cells and their function. ...Nor, indeed, do we fully understand _any_ cell. Case in point: Both neurons and glial cells feature Golgi apparatuses, which remain mysterious... As we neither understand nor can simulate a brain made up of neurons and glia, what makes you think that our lack of a nematode simulation is a big deal? Seems to me that it is rather to be expected. But when our biological models improve, simulating a nematode brain -- or _any_ brain -- will not be beyond us.