Thread: The secrets of intelligence lie within a single cell

Yes, someone posted an article that refered to his exercise of altering the signals to be close to audible speech awhile ago, and everyone called it bunk. It is a little more effective when taken in context.

I think about stuff like this when I think about extraterrestrial life. I wouldn't be surprised if there was an extraterrestrial intelligence out there that believes the Earth itself is an organism and each individual person can't possibly be intelligent on its own, especially given the incredibly predictable way we live our lives.

I forgot to mention, after reading this again:

"Yet we model this cell as little more than an organic transistor, an on/off switch.
My hunch is that the brain's power will turn out to derive from data processing within the neuron rather than activity between neurons. "

Neuroscientists most certainly don't see neurons as on/off switches, and it is well known that data processing occurs within neurons as well as between. We understand that neurons receive thousands of inputs, and it has already been shown that processing of these inputs occurs within the cell.

Originally Posted by Pensive Patrick

It's an interesting idea but completely unfounded - and seemingly 'intelligent' behaviours in cells can all be explained as part of complex molecular processes in which no sentient life is involved.

I thought the same thing. He says that molecular biology is unable to explain "intelligent" cellular behavior, but all of cellular behavior is a consequence of molecular networks. Those networks are used to compute the optimal output based on inputs coming from the environment.

Originally Posted by Pensive Patrick

I forgot to mention, after reading this again:

"Yet we model this cell as little more than an organic transistor, an on/off switch.
My hunch is that the brain's power will turn out to derive from data processing within the neuron rather than activity between neurons. "

Neuroscientists most certainly don't see neurons as on/off switches, and it is well known that data processing occurs within neurons as well as between. We understand that neurons receive thousands of inputs, and it has already been shown that processing of these inputs occurs within the cell.

I agree. That's just a problem of modelling things that happen on very different scales. You can either model things happening in one cell with high detail, or you can model networks of cells by considering individual cells as (more or less) black boxes. I don't think we have enough computational power right now to do both at the same time and we certainly don't have experimental methods that can do both. This doesn't mean that we generally think of neurons as transistors, it just means you get good enough predictions from the model by reducing the complexity and the number of parameters.