[Nuke]

ternary is easy to isolate with diodes, so its not hard to distinguish high, low and negative high. you can increase the throughput of a single wire by 50% over binary, with only a few extra diodes. thing is that computers are so damn fast that many high speed busses use differential signaling, where you have a positive signal and a complementary (negative) signal on the other line. a positive voltage is complemented by a negative voltage on the other line. ternary might not work well, as the middle level (0) has no complement. also a signal does not transition instantly. it takes time to go from 1 to zero, and for a ternary signal takes twice as long to go from 1 to -1 as it does to go from 1 to 0. you just have to allocate time for a full transition to occur before the signal is valid, of course this is costly.

[General Battuta]

DNA is a way to store data. It doesn't specifically "operate" in base 4 because the DNA replication doesn't use mathematics - it just replicates, the only logic involved is the pair forming of the four bases. DNA is a continuous data block that just happens to use four different bases (in CHEMICAL context) as its method of storing the protein encoding instructions. Making a logical operator that uses DNA would be quite hard (and would likely require use of RNA to modify the DNA sequences), but as memory, DNA would be better suited. I am unfamiliar with the proposed DNA computers' principles, so I won't say any more on that subject.

Human brain, on the other hand, is a neural net with binary signals, but much more complicated than binary logic. The signals are either on/off (electric potential opens calcium ion gates between nerve cells and electric potential pulses travel through synapses, not as electrons like in metallic conductors, but as ions more similar to what you have if you put two types of electrolytes in two glasses and connect them with a wetted paper strip, and then put anode in one glass and cathode in the other glass; you can run current through the wetted paper strip in form of ions traveling through it.

The neural network formed by nerve cells and their connections is, then, hooked to a LOT of input/output nodes, and is largely dependant on those inputs and outputs to function properly and meaningfully (sensory deprivation is a very nerve-wracking situation, literally).  And while individual nerve signals are on/off variety, the brain still interpretes sensory input as largely analog signals, depending on how many nerve endings are sending the same signal, in which case the signal amplitude increases. Each brain is has similar parts, such as the main input/output lines up to the spinal chord and brain stem, and their direct handling areas (for example visual cortex is roughly similar in structure for each person and in roughly same location, too), but each brain is also individually structured based on genetics and the experiences forming new neural pathways.


The resulting jumble of nerve signals results in personality and consciousness, but it cannot really be looked as analogous to a "computer" as we see it. There are a lot of parts in the brain that do tasks similar to "computing" - most of it subconscious routines such as breathing or hormone control, as well as balance handling which affects things like motion control, image stabilization, eye tracking, and a lot more. But even then these subroutines can't really be thought of as binary computers that get input values from senses and send output values accordingly; it's more of an analogous system that has assembled to respond to stimuli in a way that produces certain results.

It would be expected that a neural network assembled with evolutionary algorithms would become largely similar system, with no specific designed features, but instead stuff that just works as needed.

not exactly wrong but not completely right

[Herra Tohtori]

not exactly wrong but not completely right

[Flipside]

Well, I still say the real strength of computing comes from Networking, it's that which has encouraged the leap forward of digital technology far more than internal processing versatility, whether that be in the form of mobile devices, cloud computing even things like SETI online show the usefulness of it.

The problem with non-binary systems, for me, is the fact that even getting a simple binary signal to remain stable and readable across a variety of telecommunications networks is quite a difficult job. Voltages are not universal, and so a computer actually has to look for the difference between the two signals it gets rather than distinct values and, at least until recently, those systems were very unreliable for data transfer in the first place.

I see multi-state computing to be something similar to the IDE interface. The whole reason it was adopted in the first place was because it was believed that sending multiple signals at once would be faster than serial systems. This was true for a short while, but the reason that SATA was developed was because making sure all the signals arrived at the same time on an IDE port was actually turning out to take longer than a modern Serial connection.

Whilst more modern communication systems might be a bit more receptive to 3+ state signals, I still don't think we are quite there.

[General Battuta]

close enough

wellllll

nerve signals aren't always on/off digital. last time i was really involved in neurobio, the field had just begun to agree that neurons transmit both a binary action potential and the analog signal present in the cell body.

it's also important to note that this

electric potential opens calcium ion gates between nerve cells and electric potential pulses travel through synapses

is a little misleading; if a cell fires an action potential, that action potential doesn't flip adjacent cells ON or OFF - rather, it contributes an excitatory or inhibitory potential to connected neurons (EPSP or IPSP), altering the probability that that cell will fire in a graded fashion summed across all contributing nerves. so individual nerve signals aren't exactly 'on/off' variety; even the action potential itself, as it turns out, isn't on/off

this is a really important mechanistic difference between computers and brains and it's one that makes computational neuro a tricky field

the brain still interpretes sensory input as largely analog signals, depending on how many nerve endings are sending the same signal, in which case the signal amplitude increases

this is also a bit problematic. there's not a direct correlation, exactly, between number of nerves firing and signal amplitude, nor between stimulus intensity and number of nerves firing. some sensory nerves inhibit others; some are frequency-dependent; some have specialized modality-linked functions. psychophysics is also a really interesting field.

it's important to note that the brain does a VAST amount of postprocessing on most sensory data, much of it additive and interpolative rather than reductive

The resulting jumble of nerve signals results in personality and consciousness

It would be expected that a neural network assembled with evolutionary algorithms would become largely similar system, with no specific designed features, but instead stuff that just works as needed

this is...a very interesting question, and one that i'm not totally prepared to hypothesize on (regarding the convergent development of a simulated neural system)

i spent the weekend in the neuro lab at NYU and they are really ****ing with what i thought were accepted dogmas in the neurobio and neuropsych field

[Herra Tohtori]

Very interesting. Especially the part on action potential not being strictly 0/1 on individual signal basis, I have no idea how that would work. Are there multiple channels in individual synapse connecting two neurons? Or does it work with general intensity of the action potential, or a timed sequence with pulse data?


Just one thing I disagree on - the miracle thing... what I meant is that we don't exactly know how it happens (due to immense complexity of the whole process) but that the brain definitely does produce the individual personality and consciousness (however you want to define it), no miracles involved.

I just suspect we aren't really well equipped to categorize and analyze the process of how consciousness emerges from the net activity of the brain, as it's an evolved process and we're sort of geared to think on design basis. All that the brain does is based on a long ass evolutionary algorithm that has produced what we have now, there is probably a LOT of leftovers from solutions that didn't quite work but left unused structures (one example would be epithalamus and its connection to parietal eye) that might end up having a similar role as what we thought was "junk" DNA.


Strictly speaking, you could look at any current gravity theory and paste that "then a miracle occurs" image there at some point, too. The fact that a theory has gaps in it doesn't discredit the theory, only limits its field of application.

Similar to how we can empirically observe the existence of gravity, we can clearly say that things such as personality and consciousness do exist. And while we can't exactly say HOW both phenomena work, we have a pretty good idea that gravity is connected to property called mass... and that personality and consciousness are connected to neural activity in the brain.


Of course, all this depends on the definition of term "miracle" as well.

[General Battuta]

Just one thing I disagree on - the miracle thing... what I meant is that we don't exactly know how it happens (due to immense complexity of the whole process) but that the brain definitely does produce the individual personality and consciousness (however you want to define it), no miracles involved.

yes, any good scientist (or level-headed thinker) is a strong monist materialist by this point, myself included

the comic calls attention to an understandable oversimplification of an extremely thorny scientific problem we have yet to completely resolve

you probably do not need to expend lots of words convincing me that personality and consciousness are encoded in brain matter given my participation in events like the teleport threadnought

[Herra Tohtori]

It was mostly for other people's benefit.

[MP-Ryan]

Very interesting. Especially the part on action potential not being strictly 0/1 on individual signal basis, I have no idea how that would work. Are there multiple channels in individual synapse connecting two neurons? Or does it work with general intensity of the action potential, or a timed sequence with pulse data?

Action potentials involve a series of ions.  The simple explanation talks about calcium, but calcium, sodium, potassium, and a few other ions all play roles in action potential traveling down a synapse.  All of these are involved in both the net result - binary - and the magnitude - analog.  If I'm thinking of the same thing batts is, action potentials can "partially" fire, resulting in a lesser response at the synapse.  As action potentials reset incredibly quickly, I believe the synapse also responds to pulse.  If you want more detail, I'd have to go back and look it up.

[stinkyFeet]

Democritus was bound to regard the soul as material (composed of round, smooth, specially mobile atoms, identified with the fire-atoms floating in the air)

We're still at the same spot we were 2000 years ago on this front. Nobody has a clue, because it's not something we have any moral way to study in the slightest.

[General Battuta]

Democritus was bound to regard the soul as material (composed of round, smooth, specially mobile atoms, identified with the fire-atoms floating in the air)

We're still at the same spot we were 2000 years ago on this front. Nobody has a clue, because it's not something we have any moral way to study in the slightest.

well we have as much compelling evidence for the immortal soul as we do for the phlogiston, giving it a pretty solid scientific position for the time being

[Mars]

We're still at the same spot we were 2000 years ago on this front. Nobody has a clue, because it's not something we have any moral way to study in the slightest.

What color is the unicorn that you see? Mine is green!   
EDIT:
(There is just as much evidence for an invisible unicorn only I can see as there is for anyone's immortal soul. This post is not here to taunt, it is to point out the inherent error in the OPs logic)

[stinkyFeet]

Ah, sorry about that, used the wrong word, "soul". It was a quote from some website. Obviously I don't think there's some part of me that's going to Hades like that person, It was a Greek philosopher by the way, not Christian, The 2k was just rounding, one significant digit, see. I don't even know where immortal came from.

I was trying to point to some guy theorizing on the same thing back from the 2300 something years ago. Also to point out that whatever consciousness is, we haven't made any progress on identifying it. His idea was silly but also on par with anything else I've seen.

I've long since given up on trying to convince people that my unicorn is real. Btw, yeah, it is green.