A computer programmers theory of dreams

Quest for Lucidity General Lucidness
1

Last year I was studying neural networks (an ai concept) and I think it has lead me to a plausible theory of why our body needs to dream.

It starts off with the fact that when you’re born, you have absolutly no programming at all. What I mean by this is you have no knowledge and you react purely on a few basic instincts. If your hungry you cry, if you need to pee you pee.
The way that our brain works can be simulated on a computer. There are a number of input ‘neurons’ which read in data. In your body these would be touch sensitive neurons on your skin, neurons linked to your eyes ect.
At the other end of the net there are output neurons. These would be motor neurons that allow you to move, talk ect. In between are loads of random neurons that have a value (in real life it would be their conductiviy). When input neurons are fired, the signal goes through the net to the out put neurons. Then it is decided wether the out put is useful and if it is, the net is altered to promote this kind of behaviour.

In short, you put in an input, see what the output is, if you like it, you adjust your brain to act more like this. If the output is bad, it is ignored.

The problem with this kind of system is that you have to run data through the net a lot of times to make substantial difference, and you need to experience a variety of situations. This is where dreaming comes in. In the human body you cant just loop through a load of varied data. When you dream, I beleive that your mind produces a wide variety of situations to run through your net. In your dreams you may see monsters and all sorts of other things that you either haven’t experienced, or don’t experience very often. This allows your ‘neural net’ to adjust itself to situations that would not normally arise.
This behaviour is very important. Take ancient man. He may never have seen a tiger or fierce animal before. When he see’s one he wouldn’t know to kill it or run away. The thing is, he’s probably dreamt about fierce creatures, and his mind has developed a sense of fear.

If we didn’t dream, I don’t beleive our bodies would really have the ability to respond inteligently to situations we haven’t been in before.

I would appreciate any comments on this theory, even if you disagree. If bits of it don’t make sense I’ll exlain them properly if you ask.

2

Interesting theory. Actually neural networks aren’t strictly like brain neurons. Don’t ask me how but I read that once.

3

Sure they are. The concept of a neural network came from looking at how our own brains work, and trying to simulate this electronically. Realistically, it isn’t yet possible to simulate a network of the same intricacy as our minds (taking into account link degradation, chemical influence, etc) but we can still clearly see the similarities between the test simulation and existing biological demonstrations.

From an evolutionary point of view, the above theory does make sense. Dreams provide experiences unlike anything you’ve encountered in real life, and perhaps this is to prepare you for their eventual occurrence. It’s interesting to ponder.

4

I program on computers, so here’s my theory.

Yes, but it is also instinct, if you scream in a babies face it cries for its parents to defend it (or whatever).

Your brain does something (whatever it is) that requires energy replenishment. During non-REM your brain tops up its “turbocharges,” and as a requirment, glycoen stores in the brain are widley drawn upon when making the neurons work at their full capacity.

When these stores are run down they must be replenished.

To do this the brain has to enter a state in which the brain throws the neuronal membranes into a suppressed eletrical state called “hyperpolarisation” meaning the neurons become inactive and unresponive to outside stimuli. This results in our non-REM deap slow waves.

But it comes at a price, as everything does. Hyperpolarisation can only be achieved when precious Pottasium ions are released from the neurons. This sate could only be maintained for so long before the loss of ions (and therefore damage to the cells) is permenant.

Everynow and again the brain has to pump potassium ions back in, repolarising each membrane and switching the cortex on again while keeping you in sleep. This is what is experienced as REM sleep, and firing ions is achieved by brain activity (Dreaming).

Back to Non-REM. During Non-REM your brain is also slowly repeating the same activity brain patterns in real life changing and developing it slowly.

During REM it is doing this, but the dream is much more disordered due to the repolarisation going on, and neural networks being replenished with ions.

So if you wake up in Non-REM you feel sleepy because your neurons are in “hyperpolarisation” state and waking duing this state acts like a REM period, reionising your brian neuronal membranes.

5

The primary reason I disagree with this theory is because every thing in our dreams is made of something that we already know. If you have seen, heard o or thought of a big scary beast that will hurt you, you aren’t going to dream of that happening.

It is more like the interpreted result of giving random input to an already established network. This corresponds more closely physiological studies and theories on the origin of dreams. Like technodreamer said, these nuerons need to fire, and what better order than the natural order of chaos, or randomly.

6

“The primary reason I disagree with this theory is because every thing in our dreams is made of something that we already know”

I dont know about this one.How do you know?This is only because we used to think this way and if theres something “new” we easily say “oh it must have been in my subcounciousness somehow”.
Not necesarily the fact.
take care:)

7

umn… jack…

i mean to say that every thing we are is everything we know… not subcounciously or counciously… but is the sum of all the millions of seconds of events you have experienced in your life… so yeah… like i said

8

I can’t say how true this is but as sure as I can remember, I have encountered places, locations in dreams that in no way resemble any I have seen before…

Now if by saying something we have already seen you mean the mind grabbing random locations or objects and merging them into this ‘new’ creation formed of other ‘memory’ locations then I still believe it fascinating :eek:

When I find myself inside a dream, the feeling reminds me somewhat to the one you got when you where a kid and just stepped through the gates of Disney World, or where sitting on the edge of the Grand Canyon. That same excitement and curiosity…
Rarely do I visit places which I remember having seen in real life…

I admit that some of the places did resemble real-world locations. Especially basementsl or insides of apartments, these usually appear pretty unchanged in my dreams, except that maybe they’re connected all by the doors!

What is magnificent is the outside world in dreams. Unreally bright, crytal blue skys… A couple of times I’ve dreamt of this amazing plain… Long grass swaying gently in the breeze. On the edges of the horizon to my sides random groves of trees… The distances are grand canyon like! breath taking…

Straight ahead of me, there’s this great temple of a sort… It has this old, dungeons and dragons feel to it, kind of ancient and peacfull… It has a couple of towers which reach pretty high into the sky. (I remember looking up and this unusually bright sun shone behind one of the towers, awesome lens flare. I remember thinking wow! these are COOL graphics :tongue:

On the other hand for some reason interiors are always like mazes, dimly lit, and I alway have this feeling of either being lost or confused. Since each room usually leads to where it isn’t supposed to. And rarely, to say never, have I been able to reach outside from inside. (hope that wasnt confusing!). One of the times I emerged from an elevator, the doors opened and I found myself on a grassy hillside, I felt so relieved!

sweet dreams :peek:

9

Its such a pity I only came across this thread now :confused:
Unfortunately most of what I would have said, has been said :razz:

Both those theories of neural renactment, and the chemical repair/replentishment are things I’ve considered, and both have a lot of truth, I feel. I had slight differences in the way I believed in them though.

They also don’t seem to match together 100%… as both claim to be the source of the dreams, as explained here!

My personal answer:
I believe the dreams are caused by the ‘going over’ of the data, and while going over this data, the chemicals are replentished at the same time.
Thus neurons that are not considered important, are not replentished.

This is no different to repetition in the waking world, but by dreaming of different slants on known information in a dream state, you are repeating things that are important to you (or stand out), and thus replentishing the neurons important to you.

I was actually intending to attempt a neural net with the concept of chemical levels… and see what that does. Unfortunately I got distracted.

Sure, yes, you’ve seen all those parts of your dream at some point in your life. But you’ve probably not seen that exact combination before (not including dejavue and recurring dreams)
The brain works on associations both big and small, and the combination of those associations is what produced the variety of your dreams.

Thus its still based on your experience, and what you’ve learnt, but its a new slant on the existing knowledge. A different possibility or scenario to a collection of known situations.
This allows you to eg. learn how to kick the ball at its best

[EDIT]
If you follow through a conclusion on that, you’ll find that lucid dreaming in every possible dream, with absolutely 100% control, could seriously affect your natural learning abilities by imprinting the less important associations, and neglecting the more important ones.

Of course, in my opinion, this would be near impossible anyway, as LDing in every single dream… and then preventing the free flow of the dream with 100% control… is highly improbably and I would venture to say impossible.

10

Wow - logged on a year later for the first time in ages and the first thing I looked at was my own thread!!

11

welcome back :wave:

It’s good to see how this thread has slowly carry over through the year :wink:

12

Great theory but what about the idea that the Brain needs NO external stimuli to react the same as if it was awake? The Brain cannot discern between two different realities: that of the Dreamworld and the other of the Physical.

When they ran experiments and asked people to observe a Cola Can on the desk and they watched and measured which part so the brain reacted. They repeated the test with the patients eyes closed and asked them to THINK about the same Cola Can. The same exact parts of the Brain reacted the same as if the patient was looking at the Can.

This is why we rarely “realize” we’re dreaming and must train ourselves to do so. When this happens, many people describe their Lucid Dreaming experiences “more real than real.”

So what does this say about the two opposing realities that we experience? “Science” and skeptics alike agree that the Dream World is produced by the Brain and is not “real” in any way. Yet the Brain itself says that the difference between our ‘dream reality’ and ‘physical reality’ is there is none. Science is still completely up in the air how Brain activity, which consists as stated above, a Network-like arrangement of random firings of neurons, comes together to form a coherent and singular idea of “I” and “Self.” The Brain also does this at a slower rate that can account for the mere instantaeous reactions and thoughts we have. I can drive my car, talk on the phone, think about a past experience (with visualization), listen to music, drink a drink, eat, and still maintain “in the moment” enough to drive my car down the road without crashing. Not to mention all the stimuli that is coming into my brain without my perception of it (wind, temperature, sounds, sights, etc…) The standard Brain model cannot describe how it does such a task with such “slow” chemical processes.

13

yea… i belive its true… :peek:

14

I don’t agree. I can give you the exemple of sex. When I was young I was dreaming a lot about sex and I never had experienced it. When I did sex for the first time that was exactly like in the dream.

Same thing if you put on exemple when you here hurt in your dream. Its exactly the same thing even if you never had this kind of hurt.

15

I’ve been working on neural networks too and here is my theory. The main thing you have to know about the relation between neural networks and dreams is that during REM-sleep, input neurons are fired randomly. This random firing create dreams.

I agree with this. So in what could it be useful? There are two possibilities. Here is the first one: when you give random inputs to an already established network, you’ll not have random outputs. You’ll just get a copy, a map of the neural network. Hence it could be useful if you want to have a backup of the network. Thus this process can be used if you want to save data in another part of the brain. Moreover I know there is something about neural network learning. I don’t know how to explain it clearly but if you want your neural network to learn new data, it’s likely to erase previous data or give bad results. A way of avoiding this and merge previous data with new data is firing input neurons randomly. Thus it has two uses : saving data and integrate new data into the neural network.

And here is the second possibility. When learning things, a neural network can be driven mad. For instance, suppose you have to make it understand two situations A and B. But instead of mixing randomly the A and B input samples, you give it all the A samples first, then all the B samples. If you compare this to human upbringing, it’s like putting a child during many months in all the situations in which he’s punished, then the next months in all those in which he’s rewarded. You’ll easily understand that it will create a psychological problem to the poor neural network. How to correct this? Will we become mad on the first day when many bad things happen in the same time, for instance your boss tells you you’re fired and when you come home your girlfriend tells you she found another guy? In this case becoming mad would mean having the same wrong output for many different inputs.

As people don’t know how a neural network runs, let’s imagine it as a sand beach. When you pour water onto the sand (the input) it creates a little stream (the output). If you pour water on another place, it creates another stream. But if you pour water in differents places and it always create the same stream, there is a problem. Now imagine it’s raining on the sand beach (random inputs). It creates a topological map of the sand beach and you can easily see the streams. If you see that all the little streams converge towards the same big stream, it means there is a problem and you may correct it by localy modifying the surface of the beach (that is diminishing the values of some neurons). It could be another use of random firing of neurons, hence of dreams.

16

Really? :eh: Well, for me it was definately not.

Hm, that’s another funny thing. If I’m bitten by a snake in a dream, I have the right feeling, for I know it (ugh). However, if I’m shot at in a dream, it hurts, but I suppose a real shot would hurt a LOT more…

I have a question for any programmer that reads this: during my life, I have changed the way I dream. If you look at my dreamjournal now, you might notice me dreams (at least more significant ones) no longer represent unexpected situations, but more like metaphors and encenations for whatever has crossed my counscious or subcounscious mind during waking life. It feels like me trying to interpret myself. How/why did this happen?

17

I haven’t read this entire thread yet cause I don’t have time right now, but I’m going to since it looks very interesting. In there mean time Ijust wanted to say to Sepultra that while you had not experienced the actual act of sex, I’m sure you had experienced the different sensations and emotions that one feels during sex and your mind was probably just able to combine them to create the effect.

18

Do you try to interpret your dreams, Bruno? Or does it mean that you can more easily understand your dream metaphors? Anyway, I’ve no explanation.

I would just add to my previous post that I’ve watched a neurobiological lecture and that the topological map example I used to explain neural networks was very good. :grin: Neurobiologists just use quite the same, as they consider there are “attractor pools” of neurons, which width and depth vary. They are represented like topological maps. The biggest the pool, the more your ideas and dreams are likely to converge toward it, like rain into a stream or a lake. That’s what we would call a dreamsign.

19

Moreover, I’ve just read that REM sleep is said to reactivate those “attractor pools” so that it is exactly similar to what I said above in my “beach under the rain” metaphor!

/me feels he is a genius :content:
…
*but he wonders if someone has understood :uh:

20

This is a rather popular theory for dreaming, I must say! :happy:
One thing it doesn’t explain though, is why REM deprivation may result in death, if all dreams are just for ‘practice’!