What Colin is saying here comes in various "flavors". For example... 1. You could simulate individual neurons, but you would probably miss the contributions of various integrating mechanisms that accumulate important statistics. 2. In a few more decades we might know and be able to simulate all of the integrating mechanisms, but still be clueless regarding field effects. 3. With lots of physics research we could understand field effects, but still they would be difficult to simulate because each neuron may/probably be significantly and importantly influenced by millions of other neurons.
Centuries of future research will eventually not only show how all of this works from a computational perspective, but also show how to build efficient special-purpose processors to perform these computations. Will these special purpose processors be any more efficient than our own brains? Colin and I doubt it. Colin and I share broad agreement, but disagree on some details that time will eventually resolve, like I believe that the Hall effect works to make all nearby neurons mutually inhibitory, while Colin doesn't think magnetic effects are significant. In any case, no matter, at least for another century or so until we understand the other pieces of this puzzle. Some idiots think discussions like this are irrelevant, because they somehow "think" that these chemical, electrical, and magnetic effects aren't important for computation, when all evidence is that they dominate computation and show just how complex thought is. At present, progress is nil, because no one seems to be working on the vast ignorance in this area. We could get past this in a decade or few, but at the present rate of progress, we will never ever get there. This is probably good, because I see no good but LOTS of bad awaiting the creators of such systems. Steve On 10:43PM, Thu, Sep 20, 2018 Colin Hales via AGI <[email protected]> wrote: > Hi MP, > Two things regarding: > > On Thu, Sep 20, 2018 at 5:30 PM MP via AGI <[email protected]> wrote: > >> My position: >> >> Who we are is nothing more than a few billion neurons in a calcium >> enclosure and an annoyingly inefficient vehicle. >> >> So our intelligence, perceptions, actions, memories, and so on and so >> forth do indeed exist. I don’t see any reason why we can’t recreate that in >> silicon. >> >> Why we haven’t so far is another issue. To know human intelligence is to >> know ourselves, and we have so far BARELY been able to decipher the brain >> of a honeybee. What’s stopping us is from fully understanding the human >> brain and the relationship between neurological structures and intelligent >> action. >> >> >> >> *"I believe it’s possible, however. If it exists, it can be simulated. "* >> > > What if the answer to this is "Yes it can be simulated, but it's > irrelevant". What if the physics responsible for consciousness (the 1st > person perspective of being that very brain physics) supplies access to > information in a fundamental way that is > > a) gone if you simulate the physics, along with whatever aspect (say X) of > intelligence is critically dependent on it. > b) Perfectly naturally part of brain physics and right in front of us, yet > .... > c) to simulate it, you'd have to know already what went missing when you > simulated it. > > If that is the case, then you can no more simulate consciousness to get > aspect X than you can fly by simulating flight physics. Not only that, the > belief you can simulate it is actually stopping you from finding out if you > can or cannot! > > *"I don’t see any reason why we can’t recreate that in silicon. " * > Me either! We can find out the answer to the question! All we have to do > is put the brain physics on the chips instead of the physics of a computer! > Recreating it (the brain's signalling physics) in silicon is not 'putting a > computer' in that same silicon. Utterly different. In what way does it > functionally matter? Exactly. That's the whole point! Well you'll never > know if you never do it. Therein lies the rub. > > This problem is still a problem, not because we can't solve it, but > because we don't try! We are currently trapped in the grip of a religion in > the form of a delusion that 'computing a model of a thing' is literally > 'the thing', without ever actually testing it to be the case or not. To > test it you put the brain physics on the chips. We can do this! But we > don't. :-) > > And until we do, we'll still have endless posts like this year after year. > > cheers > colin > > > *Artificial General Intelligence List <https://agi.topicbox.com/latest>* > / AGI / see discussions <https://agi.topicbox.com/groups/agi> + > participants <https://agi.topicbox.com/groups/agi/members> + delivery > options <https://agi.topicbox.com/groups/agi/subscription> Permalink > <https://agi.topicbox.com/groups/agi/T2e5182d7ce6527f7-M3f6479f378ae1a0ea90db608> > ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T2e5182d7ce6527f7-M9a15b498c2773aa8713d98fc Delivery options: https://agi.topicbox.com/groups/agi/subscription
