What's the Code for the Video in My/Your Brain?

[ProJiJi]

This question came to me as of late because it seems the chemo and sleep aids I have to take have been causing some seriously hardcore dreaming on a more regular basis than in my earlier 60ish something years - - - knocking off the first ten years, as I don't know if dreaming was any sort of problem (or fun) from birth to maybe ten years of age.

And make note of the use of the vocabulary "problem" - - - as it seems many of the recent dream episodes are very unpleasant.

Okay, enough of that - - - now a lot of this video stuff discussed on this site boils down to writing code, yes?

Well, who knows the code for that "video" in our brains?

Let me take this one step further, as this may be the true origin of this crazy thought process; the chief doc for all my cancer treatments and subsequent other stuff recently had me do this 'brain wave' test thing and those tech folks stuck a whole bunch of wire end connections to my head and then did interesting things like using different light waves and such and I guess they got some chart and compared that to what is standard and for those that claim I am crazy, I have to disappoint you on that test. Seems I came out 'normal'.

Anyway, this brain wave thing isn't my imagination. And they have standards, too.

So why can't we capture a dream? Which would mean identifying the code for a video. Just that the video is in your head/brain.

If one of the younger VideoHelp Community citizens were to be able to identify that code, you'd be quite famous and rich, like many of those that wrote and copyrighted the codes that are so common on these computer machines, and the little Star Trek machines named after that robot that used to be written with a lower-case 'a' -- android.

So who of you can write the code for the videos I am watching when I'm supposed to be doing that thing called 'sleep'?

Or who knows where that research is already being done?

[El Heggunte]

....So why can't we capture a dream? Which would mean identifying the code for a video. Just that the video is in your head/brain.

Because the problem is infinitely more complex than you think.

Besides, "capturing dreams into video+audio" would be a terrible waste of resources in most cases,
considering that most people "are trash", according to that famous letter that Sigmund Freud wrote to Oskar Pfister.

[ProJiJi]

Yes, this is something that is so farfetched that in the present day it seems too complex, but I'd bet a whole bunch of money that maybe 100 years down the road (moving forward) it might not be viewed as so complex by the average folks on the planet, and off the planet. You reckon those that are "off the planet" right now view such matters as too complex?

By the way, thank you for the correction. Obviously I am no video pro, or that "audio" point would have been added to my OP.

As for that thought process about most dreams being trash, I'd have to think about that one for a bit. In the brick-and-mortar world "trash" is becoming more valuable because of technology. I use "more" because it is an age old use of "trash" that was how land was fertilized when "trash" was first used. Well, close to "first".

Interesting, too, that thought process about using "resources" because that was the very thought process that kept many "modern" devices from being provided to humans earlier than they were. And only being "invented" in a useful form because some nutcase kept at it.

One more point, if I may; I made note of unpleasant dreams and if one of you smart folks can find that video-audio code quick enough I can use that info to turn off the code of the unpleasant dreams, yes?

[_Al_]

I'd say what can help is good vibes and feeling (perhaps food and good air, which propels above). Waves/frequencies/vibes could be measured and recorded, what part of brain is activated. But our brain image cannot be recorded. Too many things and too many random things is involved.

In our brain , down to neurons and such, it is random and mess. It is all based on experiences. Our brain works different, based on what happened before. And ways were found along the way, no one has the same exact process. It cannot be printed or as in sci-fi put into computer and such, that will never happen. One stray neuron ending up on a destination point might cause totally different outcome. If someone could record your brain would mean, it could even predict a future, think about it. Sitting in a room, where nothing is changing, everything is still and quiet. If you have mapped brain, you'd have to know what will happen with you. It is not possible. Same like with a physics on a very small scale. Those devils are random. So they are in our head. We can control them in real world, making those experiences useful, but in a dream devils could dance, hard to contain, if there are good vibes present, perhaps that could be subdued.

Or the way to realize it is we could think about why we suddenly recall something. Or we do not. That is a random devil thing, triggered by merely living /flashed impulse / neuron landing somewhere at a proper time and place, sure heavily influenced by experience/situation/our senses before, but anyway. "Hey, I forgot my pancakes are on" Why would I recall that just now at this second and not a second later. Can AI suddenly recall a thing? They are going straight forward thru a code, their code has to encounter a function that checks things or better, they'd set internal timer. We set an approximate timer on in our brain but we can forget anyway. . It is an avalanche of processes and interactions simultaneously where our brain makes up shit and priorities of all of this to desperately patch things up and sort it along. I'm afraid that mess cannot be recorded.

[thecoalman]

This question came to me as of late because it seems the chemo and sleep aids I have to take have been causing some seriously hardcore dreaming on a more regular basis than in my earlier 60ish something years - - -

If I were to take a wild guess with no real medical knowledge perhaps it's not the dreams that have changed but instead your sleep is not as deep.

I know I dream but recollection of them is vague at best unless I wake up while they are happening and even then I would have to write it down immediately. Waking up for me is not common and usually because of something severe. I used to ride bikes a lot in my early and late teens, I had some of the first mountain bikes made in the early 80's when they were still glorified ten speeds. Driving off a cliff on my bike was common one to wake me up.

[TreeTops]

Here's something interesting: My wife can wake herself up, to stop a bad dream. I can't.

[Hoser Rob]

Computer information systems are all totally based on protocols. Information systems found in nature aren't based on protocols. Reducing them to a codec or whatever is way, way beyond us at this point. And I think will be for some time.

[ProJiJi]

Hoser Rob, your reference to protocols is of note. So much so that I decided to go poking around to see just what some of the professionals have been up to lately with regard to brain machines and much to my surprise I find that maybe in our last meeting my chief doc might have made a very bad mistake in addressing a question I put to him.

You see (if you care) this thread sort of started out as more of a challenge to some younger folks who are here in this community focused on visual/audio technology stuff, but just jumped into my actual life and my medical situation; which means I am thankful for the discussion causing me to quit being lazy and so checking on some of the research that we may have access to on the Net. Your reference to protocols, Hoser Rob, did exactly that.

And this one, Hoser Rob, you might very much appreciate:

As Emerson M. Pugh stated: “If our brains were simple enough for us to understand them, we'd be so simple that we couldn't.”

That comes from here:

https://pmc.ncbi.nlm.nih.gov/articles/PMC6032823/

2018 Apr 18;104(1):59–71. doi: 10.1002/cpt.1077
Brain Monitoring Devices in Neuroscience Clinical Research: The Potential of Remote Monitoring Using Sensors, Wearables, and Mobile Devices

_Al_, you made a reference to food and quality air, but it was when the relapse happened to me and I had to spend a straight 6 months as an in-patient at the hospital when the hardcore dreaming was so worrying that I started keeping notes. And I think those hospital folks dish out proper food stuff and their air systems keep the air clean. But it is obvious that once a patient starts down that relapse road and the chemo gets nastier all guesses go out the window about what is causing bad dreams and very poor sleep patterns.

Which takes me to your post, thecoalman; you nailed that one, for sure. My sleep patterns have never really recovered since the relapse treatment. I have been keeping records of what I have in the log as "awakes" and they run about every 90 minutes, to maybe every 120 minutes on a good night. I think I have had just a few times when the sleep pattern went 4 or 5 hours straight.

But when a patient starts down that chemo road, all bets are off. I doubt I fall into a 'normal' category. Plus they had to use a very special medicine for what was basically a third relapse, albeit a very strange one; some sort of membrane around my brain started to swell up and I had to take some dangerous medicine for a year for that to apparently be brought back to what I guess is thought of as normal. So I am not just mentally an abnormal type, but was also a physically abnormal type.

Frankly, after that second relapse I am lucky to be alive. Doc Abe confessed that he sort of tossed the book to keep me alive. You see we had a nasty time getting back into the hospital to start the relapse treatment because it was right when the hospital was full of the new Covid patients. Wife and I had to pull some strings to even get the waiting time to just a month and that was a close call for Doc Abe. I have the cancer that defies protocol in many cases, lymphoma. My understanding is when the relapse starts it can be quite nasty sometimes, and that one month wait was no help.

Good gosh, sorry for that long bio. Just trying to make the case that my own dream trouble isn't a good guide to work from. Yes, my dream trouble caused me to think of this idea about turning a dream into visible out-of-brain viewing, but we can't use me as a standard to work from. One thing, though, that I remember from the hospital stay was I started dreaming in series. I was getting a repeat and then it would move on to further the same 'story' and I remember having some discussions with some of the docs on the how-and-why of that. Seems some felt that was very unusual, but I don't remember if I studied any medical journals about that. Memory troubles have also been a problem as of late. Frankly, folks, I'm lucky to be alive. But I had some of that luck many, many years ago on active duty, so maybe I am just one of the humans luck favors.

Good gosh this post got way, way too long. Sorry about that.

[gll99]

On the subject of dreams to video:
Maybe there are no continuous images to transfer. If I close my eyes right now I can't hold an image in my mind for more than a few milliseconds. Even if it's an object like my car or a close family member. It just floats away. Yet at the same time I can construct a step by step narrative. ie... I see my wife open the front door then I see the hallway..next the steps to the living area and so on. My brains fills in the story even though I've only visualized 3 fleeting images. I imagine that dreams may be like that.

Outside influences on dreams:
I know for a fact that happens. I spent time in hospital too recently and interpreted and incorporated sounds into my dreams. The dreams didn't match the situation but my brain tried to create a meaningful scenario. Many times I remembered the tail end of the dream and made a conscious effort to remember and match it to what was happening around me in real life.

I need the washroom a few times a night and me looking for a bathroom in my dreams is quite common. The scene is never the same as far as I can tell but there is always something preventing me from using the bathroom when I find one. There are often tell tale signs that tell me I'm dreaming and to wake myself up.

[lordsmurf]

I've used AI to recreate images only seen in my head. It's never exact, but eerily close, sometimes even better. Detailed prompts is the key.

[ProJiJi]

That second reference to using AI to recreate a dream helped me hit on a question, but I first would like to share a bit of work I did late last night, my time [JST]. Last night was triggered by _Al_'s post above, but only after I did that much too long response just after Hoser Rob's post. Something was bugging me as I moved through yesterday's activities, and finally last night I took some time and did find what had been bugging me. It was a piece that was published by Avi Loeb earlier this year and fortunately I was able to find my copy sent to me and it is the first paragraph I wish to highlight here:

In a recent essay, I argued: “We should respect Artificial Intelligence (AI) systems with more than quadrillion connections, exceeding the number of synapses in the human brain. Unplugging from the electric outlet an AI system which exceeds the complexity of the human brain is similar to killing a person.”

One trouble that finally put me to sleep was I'd moved that piece he'd written out of the email account where it was originally sent earlier this year and apparently lost the link. So I have to go find that link.

But then when I came here to post his thought process that I wanted to introduce to _Al_, I then saw your post, lordsmurf, and it triggered an odd question. I'll just quote myself by copying the hard copy note I wrote a short time ago in my notebook I'm keeping for this thread:

We humans are different from AI because of something we call ‘intelligence’ and so I have to explore (ask) whether it is intelligence that creates a dream?

[lordsmurf]

AI is just a really smart parrot, regurgitating vast amounts of data, incapable of truly unique and original thought. Of course, the same can be said about many human, though not all. Focus on the intelligent humans, not the drecks.

AI is just a tool, like a hammer. Not a living pet, not a friend, not family.

[thecoalman]

AI is just a really smart parrot, regurgitating vast amounts of data, incapable of truly unique and original thought.

It doesn't simply regurgitate information. One of the key things it can do is see deep patterns. If for example you feed it every medical paper from that data it can extract unique information no human could ever expect to accomplish.

If you want an example of it doing something unique it's been solving some long standing mathematical problems.

https://www.youtube.com/watch?v=fDAPJ7rvcUw

In this case by itself without any knowledge of it determines a long standing method for doing a calculation and then goes onto discover new and faster methods.

[_Al_]

That video is an example how AI manages things as oppose to a human. First of all, models need to be learned, but humans do that all the time and even it takes painstakingly longer, 20 years and schools etc. AI tries to perfect those calculations in a sense they start to introduce a guesstimate costs and making shortcuts that are good enough. Like those tensors.

We humans operate like that already, we can decide instantly what approximation scale we would operate on. What is good enough. Actually that is even preference, not those exact comparisons and details (that is engineer/scientist/mathematician sphere only) We see someone in a distance in a shirt. We do not know exact pattern, we do not need to know exact pattern to the detail, all we remember it was a checker shirt an we remember color and size patterns. We cannot replicate it exactly but somehow we know and can compare it as a whole. If we have to compare or find that shirt we just pull it from a pile of shirts with no time delay at all, right away. And it also took just a second to remember that pattern.

That video was about (I think I understood it that way) how to implement or starting implementing similar scenarios to AI using matrixes/tensors.
To tackle information gathering in a straight line, even with a tremendous speed, is not a future of AI, it will have to constantly work in terms of humans like guesstimation, comparing cost to get that information, or just working on exactly down to floating point, where price/time spending for that could be monumentally higher. Like trying to render exact shirt pattern from a video and comparing it exactly with other patterns.

There was a question above how we differ with AI, this is one of those so far.

[_Al_]

Another AI human difference example is my presence here. I can speak in broken English with a wrong spelling (but because of spelling checker not that often) or mainly with wrong English word order. But my brain says - "It is good enough!" No perfection needed. And it could even serve as my character.

[thecoalman]

That's impressive for "broken English". You could give English lessons to some English speakers, usually the same idiots that will criticize bad English when used as a second language.

If I recall correctly years ago Baldrick mentioned he couldn't speak English, I suppose that is common these days. I'm a moderator on phpbb.com and at least half of the team do not speak English natively. It's English only and I'm always impressed in how well they communicate.

Personally I can't write or speak any other language. My grammar and spelling skills aren't that great but I'll put my vocabulary up against anyone.

[_Al_]

Thanks, that is too much praise though. I mentioned it because that is a part of a human character, a recognizable feature.

How we are going to make those AI bastards apart from each other or even make them apart from us humans? They can even fake characters and pretend, that is a core of their danger maybe. And worst of it, just mentioning it here, I might get on their list soon.

So far, we can ask them, "What sweater I had yesterday, while I was talking to you? And give it 0.5 seconds to answer while presenting 5 sweater pictures. . But soon they figure out that approximation complexity in time and humans are toast. Or we have to be really working deep to come up with questions that humas can answer in no time but AI cannot.

[thecoalman]

So far, we can ask them, "What sweater I had yesterday, while I was talking to you? And give it 0.5 seconds to answer while presenting 5 sweater pictures. .

It's far better than that.

https://www.youtube.com/watch?v=Sq1QZB5baNw

The danger down the road is keeping control of it.

[lordsmurf]

AI is just a really smart parrot, regurgitating vast amounts of data, incapable of truly unique and original thought.

It doesn't simply regurgitate information.

I fully disagree.

Look at your recent video. It put up the plates. It never thought to make a game, or have fun with putting them up. It just decided that completing the rack (puzzle) was the correct answer. It has no imagination, it simply extrapolates known activities.

If you give a command like "have fun with it" or "surprise me", you get error messages.

[_Al_]

It is definitely amazing.

The objects are fine defined there, but I'm afraid those shirt patterns and instant comparisons would not be that easy yet.

But that robot we saw, that is not a threat to people, I mean an masse, to a person maybe, same as human interaction. But it would be the system behind it, a system that controls it. If that robot is isolated and learning alone, that might be ok. Problem is , someone constantly updates it, sways it to this or that etc. Higher control entity would be a danger.

[ProJiJi]

About the idea offered, lordsmurf, of a robot type not being able to respond to something like "surprise me" or "have fun with it" I suspect we can find humans that are not able to respond to such a request, because it is not in their nature. Humans that are like super serious all the time and really don't relate to such things like having fun.

True, not many humans are like that, but I reckon they are out there and are not the type of human that can be designated as not being able to fully function in the mental aspect of humanness, like being diagnosed as a sort of - - - what is the polite style - - - mentally deficient.

Haven't we read about the human labeled as a workaholic and has no sense of humor and takes no pleasure in entertainment type things in life and has a strict routine in life that centers completely around a single purpose; if you were to ask that type to surprise you, you may get a blank stare. But behind that blank stare is a very smart human that is super good at whatever is the center of their existence, like their work.

Maybe the difference between the AI and the human is the ability to have preferences.

Have any of you read, or heard, about an AI that can like or dislike something?

What about that human thing called 'intuition'?

One other thing that popped into my head as I have been reading these excellent posts by y'all; why do we even modify the word 'intelligence' with 'artificial'? How can intelligence be artificial? Of course, I should have checked what's on the Net before asking that here, but still . . . Anyway, the question just popped into my head a few moments or so ago. And I write that the posts are "excellent" because they require more than one reading.

By the way, I could benefit greatly from a robot that knows how to properly organize my living quarters. Oh yes, and can cook. Can taste the food first. My medical situation makes it best I live mostly by myself, so this house is quite the mess. Wife's house over the hill is quite the neat style. Super neat.

Yep, I need a very special robot. I wonder if robots can do comedy? They can dance, right? We have robots that can fly our fixed wing aircraft, but I don't think I have heard or read about a robot flying a chopper. Chopper is a whole different ball game when flying.

EDIT: I just thought of an idea somewhat related to the flying; have any of you read or heard about a robot that can ride a motorcycle?

[thecoalman]

I would suggest regurgitating something is a what a computational computer does. e.g. Deep Blue beats Kasparov in the 90's, Kasparov can't calculate millions of moves but he can make intelligent ones. The computer is just simply making moves based on it's programing and it was able to beat Kasparov because of it's overwhelming computational power.

An AI chess computer is completely different, it teaches itself how to play chess simply by playing itself repeatedly and it gets stronger each game. A computational computer is doing 5 million calculations per move and the AI is doing 30K. It's picking and choosing what calculations to make just like Kasparov would but this is combined with computational power.

If you give a command like "have fun with it" or "surprise me", you get error messages.

I have been using AI quite a bit recently to improve my coding, in particular regex's. I used a dry joke which it didn't identify until I prompted it and then went onto explain my joke. Fun is like feelings, it's part of human nature. here is no reason AI can't learn what humans perceive to be fun but that's subjective. When AI gets to the point of having fun we have a problem. LOL

[thecoalman]

One other thing that popped into my head as I have been reading these excellent posts by y'all; why do we even modify the word 'intelligence' with 'artificial'? How can intelligence be artificial?

Intelligence is something humans take for granted. You can ask a five year old "What is the color of Dorothy's shoes?" and as long as they have seen the movie they can answer the question. This may seem trivial but it's not, it requires intelligence. Back in the 50's computer engineers thought they were going to be able to do this in short time, it only took 70 years. The computer that beat Kasparov in the 90's did so because it can crunch numbers which is something computers excel at but a fly has more intelligence.

There is no line of code to indicate to AI Dorothy's shoes are Ruby Red. It has to figure that out on it's own through reasoning and logic just like a human, that requires "artificial intelligence".

Yep, I need a very special robot. I wonder if robots can do comedy? They can dance, right? We have robots that can fly our fixed wing aircraft, but I don't think I have heard or read about a robot flying a chopper. Chopper is a whole different ball game when flying.

Commercial planes have been pretty much been able to take off, fly and land themselves for decades. This does not require AI. AI is only required when you need decision process. Do we fly around the storm or through the storm?

[ProJiJi]

It's true that I have been grounded for well over a decade, but I still don't think we have FARs that allow an auto-pilot to rotate (take-off) a manned fixed wing. My military flying was even further back in years, but I think the same might be true with military birds. True, they have those drone thingies and no human is on any kind of flight deck on the drone, but with birds where humans run the show I am not so sure that even in these modern times flight crews are allowed to use full auto-pilot to get the bird off the ground.

[thecoalman]

It's true that I have been grounded for well over a decade, but I still don't think we have FARs that allow an auto-pilot to rotate (take-off) a manned fixed wing.

What the regulations allow and what is possible are two different things. Self landing was introduced with the Tristar in 1972, I'm fairly confident they figured out how to automatically take off too. None of this would require AI, it just needs valid data.

[El Heggunte]

AI is just a really smart parrot, regurgitating vast amounts of data, incapable of truly unique and original thought. Of course, the same can be said about many human, though not all. Focus on the intelligent humans, not the drecks.

AI is just a tool, like a hammer. Not a living pet, not a friend, not family.

You are correct, as usual.

But the current status of the A.I. hype is absolutely insane and irrational.

SOURCE: https://bsky.app/profile/annleckie.com/post/3lecgtzec5k2v

[ProJiJi]

That reference to energy consumption and the influence of AI on said is valid. It seems that the increased use of AI is having a significant impact on energy demands. And I seem to remember that Microsoft has made some offer to get Three Mile Island's infamous power plant back on the electric grid. Well, I don't know if they will hook it to the grid, or use all the output for itself, in some manner.

Still, first time I learned how AI was increasing energy demands I was quite surprised.

[thecoalman]

Because of the natural gas boom here in the US which started around 2008 it is relatively cheap, the plants are cheaper to build maintain than nukes or coal and most important of all they are more adaptable to the up and down demands caused by renewable energy. It's replacing nuclear and coal. Coal and nuclear plants are in the same boat in the market today. They can only provide cheap electricity when operated as base load plants where they are running at max capacity 24/7/365 except for maintenance downtime. The demand for contracted guaranteed power is not what it was

Microsoft needs reliable energy and since they will be buying all the power produced a nuclear plant is suddenly the best option both for cost and most important to them, reliability.

Still, first time I learned how AI was increasing energy demands I was quite surprised.

It's more than AI, there is at last one case I'm aware of where bit coin miners purchased a coal plant.

[ProJiJi]

That is a very good reference to the "... up and down demands ..." as that is an aspect of renewable that is a big headache. Battery storage is simply not yet able to handle some demands when the Sun is around the corner helping out our distant neighbors on the other side of our globe [now we need that flat Earth always facing our good buddy, the Sun] or the winds are taking a rest.

But to get back to that AI stuff, please allow me to point you to a piece recently published by the Kleinman Center for Energy Policy:

https://kleinmanenergy.upenn.edu/commentary/podcast/why-ai-consumes-so-much-energy-and...done-about-it/

< < < And a snippet > > >

Artificial Intelligence is taking off. In just under two years since the introduction of Chat GPT, the first popular AI chatbot, the global number of AI bot users has grown to one and a half billion. Yet, for the U.S. electricity grid, AI’s dramatic growth could not have come at a more challenging time. AI is energy-intensive, and its expansion is putting additional strain on an already burdened grid that’s struggling to keep pace with rising electricity demand in many regions. In addition, AI’s energy demands complicate efforts to decarbonize the grid as more electricity–generated with a mixture of carbon-free and fossil fuels–is required to support its growth.

[thecoalman]

The irony about battery storage is that it's more applicable to conventional plants. With a very small amount of battery storage coal, nuclear and gas plants can be run as base load when they are the most efficient and you are fully utilizing the capital investment. There is an even bigger savings because you don't have to build and maintain as many intermediate and peaking plants.

The question I always pose to people is how much renewable capacity and storage is going to be required to meet the demands of 0 degree weather in the northeast in January, the sun is near it's lowest point in the sky, it's cloudy, no wind, it's going to be like that for a week and record demand is occurring at 7AM?

The capacity and storage requirements are insurmountable and you will continue to pay for conventional "backup". The cost per kWh from the conventional power necessarily rises because of under utilization of the capital investment.