Computer chips are printed on a single scale. That is a very small scale is beside the point, because it is not so small that their is any major danger of a misfire of current, a mistransmition of a bit (on or off). Our neurons, by contrast, are connected by fibers that, thought already microscopic, branch fractally into smaller and smaller and scales, and at the synapses, transmit by chemical reaction. This system seems prone to error, and indeed, mistakes and losses of memory are nothing new to human or animal experience. I, however, being the Bible-believing oddball that I am, don't believe that God makes mistakes. So I believer their must be some advantage in fractal neuro-pathways. So what is it? Here's an idea: The smaller in scale you go, the closer you are to the quantum level, the more heisenberg's uncertainty principle impacts the system, the more likely a synapse is to "misfire" a bit (on or off, yes or no), and the less the misfire matters. The misfire creates a new bit, a change in the information. Now if this bit is not compatible with the task the brain is trying to complete, it is erased or ignored, stopped at that smallest scale. If however (and this is the good part!) the new bit turns out to be useful, it would be carried into the level up, where it is evaluated in light of still more information, and this continues through the branches between and inside the neurons until only that information which fits with reality and aides with solving a problem accumulates. If this sounds like a slow way to accumulate information, think of it this way: suppose each neuron only produced one (1) bit of information every ten minutes on average. With the millions of neurons in your brain, and a couple of hours, and the information you already know, it wouldn't take you long to think of something no computer would have thought of. Our brains (might) percolate information out of randomness! Now, suppose you applied this concept to computers, replaced each circuit branch with a smaller scale network, working down as close as possible to the quantum level. The same thing theoretically should happen: Only that information compatible with running the input program would percolate through into the output. If nothing else, it would make for some tough PC game AI. Any thoughts on this theory or its applications?
I think you look at it from the wrong angle. To discuss the brain, you hardly even need quantum physics. Of course, in the end they ARE involved in there but the brain rather works by diffusing signals. Diffusing is rather a thermodynamic thing. And thermodynamic effects are what destroy coherence which is necessary for quantum physics to apply.
Rather, you should think of it as how this whole system came about. The first neuron-like cells were pretty bad at their tasks. Likely, there wasn't even one unified standard right away. You can even see that in us, where certain rather muscle-like cells in our heart, not related to the nervous system, DO transmit electrical signals in a similar way. They aren't as fast as neurons but their advantage is that they are (even) more error-resistant.
http://en.wikipedia.org/wiki/Electrical_conduction_system_of_the_heartNeuronic structures basically evolved for sharing information between specialized regions. At the beginning it was an unstructured, fairly messy, diffuse mesh going fairly uniformly all over the place in a body - that's still the case in, say, jellyfish.
The system never was designed to be perfect right away. It was only meant to save your life.
It was capable of "telling" the body to AVOID THIS DIRECTION, PRETADOR AHEAD - now avoiding that direction is fairly easy. You just have to do some random movement where the net-momentum points in any direction but that one. Not much thought or control necessary.
Or YUM FOOD AHEAD - same deal.
Slowly, things became more structured and as a result, able to accomplish more complex tasks. Resources were used more efficiently which in turns allowed for more complex structures.
Very recently in evolution, humans happened. They evolved certain parts of the brain that give rise to functionalities far beyond what's been found in any other species (although findings keep showing up that suggest, that we aren't quite AS unique as we would have thought initially. Certain high-level cognitive tasks were managed to be pulled off by other mammals like whales, domesticated animals or other grand apes, as well as certain birds, especially crow-kinds and parrot-kinds and, as far as I know the only exception out of that field, squids.
Still, the main purpose of our brain is to make sure we survive. However, most survival tasks don't need complicated solutions. Some fairly simple, most of the time easy to accomplish constraints have to be posed: Find food, sleep a bit, avoid dangers, etc.
At some point, the brain got SO good at that, that suddenly a lot of sparetime allowed for more advanced abilities like becoming social (which in turn gave EVEN MORE survivability and as a result EVEN MORE spare tmie), having a culture (There are male birds of some species that apparently waste hours of their time just to decorate their nests. The most beautiful ones are picked for mating. - that this is even possible, proves that their basic brain is by now easily capable of giving a pretty solid chance for surviving the next day without all that much effort. It's neuronal luxury) and eventually forming deeper and deeper emotions aswell as giving more and more insights to the fundamentals of logics and information itself.
However, evolution is a very strict net-benefit calculator. If you improve a thing, you gotta do so inside the constraints of your resources. So ANY biological process is ALWAYS just improved up to a certain point. Unless a serious "technical" jump happens that allows to use your resources WAY more efficiently all of a sudden and thus makes you able to waste more of those resources on expensive luxury, certain features will reach an equilibrium.
This means that, as long as the system is fail-save enough, you don't need to get rid of all the errata. In fact, a bit of noise even is within the spirit of evolution. It gives rise for new ideas that cannot be considered by a purely deterministic system.
Also, if you tried to drive your inner wirings to absolute error-free perfection, (obviously I'm simplifying things here since you yourself can hardly change what was given to you) you would suddenly hit a resource wall. So you make sure, every single bit of information your neurons process is ALWAYS correct. There is no bitflips, no bit swaps, no noise.
But to get that, insane amounts of resources would have to be wasted to get there. It would simply not work. It's extremely uneconomic.
The very same deal is true for the DNA, by the way. Mutations happen on a fairly low rate. But the main reason they happen is that it would be too much evolutionary work to get the copies better. Make the DNA any better in its reproduction step and your offsprings will actually become weaker over time. Not because of faulty copies but rather because they can't cope with the evolutionary jumps, all other species are making in the mean time.
Making it more perfect than it currently is, kills you in the long run, as weird as that might seem.
As of applications to computing: Well, a lot of evolutionary algorithms, gradient descent algorithms or neuronal networks mimic evolution or thought process in various ways. That's the huge and exponentially expanding field of artifical intelligence. Big accomplishments are made in that field almost every day. Most of which you'll probably never even hear of.
Things you can consider watching. Those might not seem related but they actually are:
This playlist gives you an introduction to Economics:
http://www.youtube.com/playlist?list=PLAEA5E9ACA1508F92Note, that evolution can be thought of as a highly economically motivated process. If you understand the concepts given in that playlist, you are likely to better understand evolution aswell.
But you don't even have to watch all those videos if they are too much. If you watch the first third or so, I think you should probably be seeing the connections. Though he doesn't ever mention these connections since he tries to teach other things in there.
This playlist is on human behavioral biology:
http://www.youtube.com/playlist?list=PL848F2368C90DDC3DThose videos are longer but I'd really suggest you to watch them. They are really awesome and clearly show how a lot of things happened that define us now.
it should work with REAL neurons and synapses to really scare the hell out of me
This is already done. It's obviously in the beginning of research but more and more successes are also made in THAT field.
Example from August '08:
http://www.wired.com/wiredscience/2008/08/how-to-train-yo/ .. probably possible, but consider large scale electron pathways need longer times, hence the reason to make em as small as possible to reduce distances, and in fact you have kilometers of electronic pathways built in a single cpu core ...
Well, if you look at the brain's structure, you'll find that the fractal organization actually keeps the average axion length very low. You have areas of minimal distance, neuronic regions that do exactly ONE task and then you have some longer connections that give rise to diffusive thinking across tasks.
Some qualitative statistics on that:
The average female has an a bit bigger global to local axion ratio than a male. That's the main explanation why they, on average, aren't as good at logics but often better at creativity. However, to be exact, both logics and creativity need both focused and diffusive thinking. Just the simplest incarnation of each tend to rely more heavily on one of the sides, respectively. Beyond simple tests, those differences become more and more vague.
Authistic people are found to have roughly the same amount of neurons as non-authistic ones. But this global to local axion ratio is significantly shorter. What you find is exaclty their behavior: The way more shorter connections allow them to have extreme focus on singular tasks. They can do those unexpectedly well. But they can't jump around with their thoughts nearly as much. Diffusive thinking is something they are very bad at.
the simplest explanation is the preferred one.
This isn't entirely wrong. But it's actually the simples explanation that has hope of development. E.g. it should be falsifiable and it should predict things.
Also, I'd like to think of god or some related entity/entities not as an alternate concept but as an additional one. Current science mostly asks entirely different questions than theology. The fields are mostly unrelated.
It's true, that quantum physics and the theory of relativity have implications that go so much beyond what we would have thought, that it's quite charming to think of all the beautiful equations as "set in stone" by some higher being.
However, that those theories REQUIRE such an entity is far from reality. Though, to make that clear, NOR do those theories REQUIRE that there is no such entity.
Create a set of rules that control an existence that may then grow/evolve
This is exactly what it's all about. Nice.
a photon has no "size" in the mass dimension, an electron a very small size and a proton and neutron considerably more etc.
Since recently, the higgs was found, I got a lot more insight on that by all those nice videos that started popping up on YT. I especially recomment watching the three video series from Minute Physics on that topic. (Or at least the two of three videos that are currently availabe)
<![CDATA[<a href="http://www.youtube.com/v/9Uh5mTxRQcg&rel=1&fs=1&hd=1" target="_blank">http://www.youtube.com/v/9Uh5mTxRQcg&rel=1&fs=1&hd=1</a>]]>As you'll see in video two (you get there from video one), mass is basically happening, at least on a mathematical, quantum physical level, because those particles bounce around a lot in the higgs field. And the reason it was so difficult to seprate the higgs boson (if what they found actually is that), is because that boson itself interacts so strongly with the field. You have to excite the field to quite high amounts (at least high for particle physics - on human world levels, that value still is tiny) to seperate this particle.
these are thought experiments, and the universe really doesn't give a flip what you think about it.
Of course, a lot of it IS thought experiments. But thought experiments, supported by data from actual PHYSICAL experiments. A lot of data.
The part where we still are entirely in a thought-experiment realm is called string-theory, m-theory, supergravity, etc. - A lot of theories with similar basic ideas, some of which more general than others. (Of the given examples, M-theory is the most generic one because it contains all five flavors of string-theory and supergravity)
THERE we do not yet have any evidence. Mostly because it's really difficult to think of experiments that can be done on earth or in space that is close to earth.
Most thought experiments on that sector involve black holes or other circumstances we can simply not rely on when being anywhere near earth.
However, I've recently seen a paper somewhere that suggested an experiment that is actually executable on planet earth. Can't exactly recall what it was but in that case, if I recall, the main problem was a proper detector. Effects caused by string theory that are not predicted by the standard model or general relativity, are prone to be very weak in circumstances that can be achieved on earth. In fact directly here we can mostly validate quantum physics and special theory of relativity. General theory of relativity right now can only be shown by observing space. - However, apparently we already found quite usable evidence for the Lens-Thirring effect, a very weak relativistic effect in which a rotating massive body causes the space around it to twist a little, directly on the earth's surface.
So in short: As long as you don't go for unifying Quantummechanics and general relativistic mechanics (note, if you exclude gravity and just go for special relativity, such a unification IS possible), you are far away from realms that are merely "just thought experiments".
Some people think evolution supports the notion that life is ubiquitous and anywhere the conditions for life exist we should expect to find it (we are not special), when actually the opposite is true. Evolution supports the notion of a single "magical" act of creation that has never repeated in 3.7 billion years. We are special!
Nope.
The reason, evolution makes it seem that life appeared magically out of nothing is, that evolution makes a very basic assumption:
Given life, life will evolve.
You need life for evolution to happen.
A description of how life came to be is NOT included in theory of evolution.
Evolution is only telling you how life became what it is today.
The field that does tell you how life itself came to be is called abiogenesis. That field is by far not as well developed as evolution but there HAVE been results:
- The miller-urey experiment and variants of it clearly proved that it's possible for reasonably complex organic structures to be formed randomly. Such experiments have been executed for mere weeks but they model the early conditions of earth's atmosphere, etc.
In various different arrangements, all kinds of biologically used amino-acids have been formed, as well as a lot of unused ones.
The substances from the original miller-urey experiment are STILL meassured. As newer technologies emerge, we find more and more organic compounds that simply weren't noticable with previous technology because of lack of sensitivity or even lack of reactivity to the substance all together. Every year, more and more substances are found in the material that resulted from that one experiment. And that was a few weeks worth of experimentation. Scale it all up, give it way more space, way more time and probably additional physical catalysts (the original experiment considered electric charge. Other experiments used UV radiation or the forming of ice.), you can clearly see how early earth soon was quite a soup of organic materials that eventually turned into bubbly foam, storing materials, becomming progressively better at storing those materials, starting to encode properties, possibly via RNA or maybe some proto-DNA like TNA, and then kicked off evolution.
- Asteroids are huge chunks of ice. There are evidences that early earth was hit by quite a lot of those, giving the planet a lot of its valuable water resources.
Probes from Asteroid trails showed that those huge ice-blocks already contain a lot of organic materials of varying complexity on their own right. So probably, the delivery of organic materials via asteroids even accellerated things a bit.
Epigenetics tells us that every choice we make from the food we eat to the level of stress in our lives has an effect on inheritable gene expression. We are choosing our mutations.
I suggest you too, to watch those lectures on behavioral biology. Epigenetics is an important part of that.
What you said there isn't entirely wrong. In fact MOST of our DNA is Epigenetic structure. E.g. What actually encodes features is about, if I recall, 30% of the deal. Some stuff is viral DNA that we got immune against. That's another couple of %-points.
But mostly, human DNA - and most DNA of highly developed creatures for that matter, consists of what's basically switches and knobs, finetuning the workings of our DNA.
Epigenetics has nothing to say on mutation. Those structures are subject to mutation in a very similar way as Genetics themselves are.
What epigenetics do is, sometimes permanently, activate or deactivate a part of the DNA.
For instance, if you are pregnant and in the last third of pregnancy, suddenly getting almost no food is a very strong signal to the baby. It will start producing certain enzymes that plug in on very specific parts of the DNA. Those sections are controlling digestive systems and such. If this happens to you as an unborn, you tune your body to be as efficient as possible with digesting and to be VERY careful with spending resources. The results? You are much more likely to become obese or diabetic.
Epigenetics is also where the famous but usually misinterpreted grandmother effect kicks in: Epigenetic features are not given to the next generation by males. Sperm just can't carry that information. But females have some of the information stored in their eggs. Strong epigenetic effects caused by such sudden hunger periods (happened historically in, I think it was holland, during WWII) or by stress (those two are the ones that are best understood but there are numerous other ones and it's very much subject of ongoing research), such strong effects can be shown up to five generations after the initial cause. So that's the grandmother effect. Your great-great-grandmother along the female line could have an effect on how you are.
If it chooses to stay in that environment the chances of its offspring mutating are greater than if it had moved.
This isn't wrong but it's not epigenetics.
Also it's always a bit shaky to talk about choosing. There isn't much choice if the only thing you can do is following your instincts as adequately as your physical body allows you. A "healty" amoebea will either already be resistant against that poison, thus it's not poison for it at all, or it will move away. No actual choice involved.
yes I for on think every cell that is living has some inherent quantum property's due to it's complex chemistry (think amplification of stochastic chaos) these quantum property's add up to freewill in some organisms but also help guide evolution over long periods of time
It's already shown that some levels of cellular mechanics only work by assuming quantum physics. However, for the cell as a whole, most processes are accountable for with classical physics and thermodynamics.
It's all a matter of how closely you're trying to look. Living cells usually are still too big to rely on quantum mechanics. So they are unlikely to get advantages out of those processes.
However, never say never. Weirder things have happened.
October 02, 2006, 09:19:59 PM
Dilber MC Theme by HarzeM