I just had a retarded idea

[Atreus21]

I had a death in the family recently, and I was thinking about it when the following occurred to me.

Suppose we created a computer, so vast in capability, that it could be connected to a person's brain from the moment of birth (or even earlier) to the moment of death. It's purpose would be to map out and catalog every thought, impulse, motive, action, etc. that ever occurred in the brain of the subject through his or her entire lifespan, for the purpose of creating a psychological and mental profile of the subject, so that after he or she died, the subject's loved ones could consult with the computer, which would create a physical manifestation of the subject's face, personality, and overall demeanor. They could ask it questions, and the computer could construct a response based on a lifetime of data collected, and therefore find the most likely answer.

Could this be called a sheerly practical if not actual immortality? Has this ever been conceived or worse, attempted? How accurate of a profile could a computer construct with a lifetime of data to draw on? Frankly, it kind of sounds like something out of Futurama, with the talking heads in glass cases.

I didn't know where else to mention this idea, so I figured here would be good.

 

[CycloWizard]

It all depends on the software. The hardware to monitor a lot of these things already exists, but understanding what the signals mean (and, even trickier, placing them in context) is something that would likely have to be done in software. The complexity of attempting that makes my head want to explode.

 

[Atreus21]

Originally posted by: CycloWizard
It all depends on the software. The hardware to monitor a lot of these things already exists, but understanding what the signals mean (and, even trickier, placing them in context) is something that would likely have to be done in software. The complexity of attempting that makes my head want to explode.

That's kind of the fun of it. Such a feat would require, at a bare minimum, mind-reading technology.

 

[reallyscrued]

Stage 3: Bargaining.







P.S. If this was in poor taste, I'll retract my message.

 

[esun]

Hmm, well I know that in epilepsy research, brainwaves are monitored at various points in the brain and can be recorded for several weeks in order to determine the source of seizures.

They produce about 2 TB of data per week for each patient. However, keep in mind that this is recording at maybe a dozen points in the brain, not the entire structure. If we were to record extensively enough to do what you're thinking of, we'd be looking at probably billions of times that much data in the same amount of time. Suffice it to say that it isn't technologically feasible at this point (storage aside, just recording the data would be infeasible at this point due to the number of electrodes required).

From a philosophical standpoint, you can call it whatever you want. I know philosophers like to argue extensively over terminology (e.g., can this be called immortality?). However, in the end, we can just say what it is: a very accurate profile of a person based on the entirety of that person's life. Whatever you want to call that is what it is.

 

[BladeVenom]

In about 20 years we will probably have computers capable of simulating a human brain. Now to get the information in there may be a little harder. Maybe it could be possible to have scanners that will be able to analyze and map out every neuron by then. Of course would anyone want that? It would be very creepy to have a computer replication of a dead relative.

I wouldn't call it immortality. After all the original person is dead. It's just a simulation.

 

[sportage]

I doubt this could ever happen.
Computers use logic. A program to simulate human illogical
decision would still have to fall under logic.
The human mind has the ability to make totally illogical
decisions. I doubt a computer program could simulate
illogic unless it was programmed to make random illogical
decisions, and that would be logic for the computer.

 

[f95toli]

This idea -or variations of it- has been around for a long time, see e.g. "Altered Carbon" by Richard Morgan.

I doubt this could ever happen.
Computers use logic. A program to simulate human illogical
decision would still have to fall under logic.
The human mind has the ability to make totally illogical
decisions. I doubt a computer program could simulate
illogic unless it was programmed to make random illogical
decisions, and that would be logic for the computer.

That argument doesn't quite work. We know we can -at least in principle- use computers to simulate any physical system, and since the brain is (as far as we know) a physical system it follows that it should -again, in principle- be possible to simulate it.
Also, I think you use the word "logic" out of context; whether or not a decision "makes sense" has nothing to do with whether or not can be simulated by a formal system (a Turing machine).
You can quite easily make a computer program "unpredictable" by simply using a random number generator (which -if you don't think algorithmic generators are unpredictable enough- can be implemented hardware is such a way that they are "truly" random); it is very likely that that such "random" events (effectivly noise) are very important for how our brains work.

 

[Shadow Conception]

The Final Cut

 

[Modelworks]

It goes back to the question of what makes a person, human ? Where is the so called 'soul' stored in the body ? Or how much of a person can you replace before they stop being that person ? I'm actually working on a animation job that is about these same questions. It mainly deals with cyborgs and how much you can replace and still keep the humanity . I saw a program on , think it was on discovery health, Alzheimer's and how memory is stored. They said that every time we access a memory we change it slightly because of how the brain stores memories.

 

[bsobel]

Go google 'the singularity'

 

[koshling]

http://timstretton.blogspot.co...-should-i-read_09.html

 

[JTsyo]

Not only would the computer need to monitor the brain but also the environment. I guess once you map the brain you can get the environment from the person's senses. It would be kind of how they teach prosthetics which nerves do what. Thing with people though is that responses change with time. So if you use data from when they were five and compare to their response when they are 40, of course the answer will be different. So I don't know that there would need to be all that back data. Maybe 6 months to a year might be enough, as long as the memories came with it.

 

[devione]

While memories are a big part of one's personality, there are other factors that make us who we are. Perhaps those other factors are impossible to replicated on a computer.

 

[Eureka]

The thing is, how different is the brain from a really really complex computer? Our 'soul' could merely be not more than memories in conjunction with computation and processing methods.

Of course, you'd have to code it in python.

import soul

 

[gsellis]

Originally posted by: sportage
I doubt this could ever happen.
Computers use logic. A program to simulate human illogical
decision would still have to fall under logic.
The human mind has the ability to make totally illogical
decisions. I doubt a computer program could simulate
illogic unless it was programmed to make random illogical
decisions, and that would be logic for the computer.

Actually, the computer only makes programmed decisions. Logic has nothing to do with it. The computer can work quite like the human brain and make equally flawed decisions, just much faster. It is just more consistant.

The problem is that you have a massively parallel analog device that you want to record. That device continues to modify its "applications" at will with new code. And the analog data cannot be quantified into set field lengths.

Example:

How do you store a smell? And then set a key to the smell that associates it with actions or then cross links it a whole other section of data that may not even be related. Just a fresh baked batch of chocolate cookies could be linked to various remembered and 'locked' memories at home. Then cross-linked to "grandma's". And the next time you smell the same, it may unlock a memory that really was linked, but the action changes. So, one time, you are walking down the street and smell cookies at the bakery. Depending on other environment variables, what memories it triggers, and current 'problem set' running, you may go into the bakery.

I would guess that even a century is not enough time to create anything that could map out the gathering tools or the decision ability in a computer to map a person.

 

[RadnorHarkonnen]

Actually, the computer only makes programmed decisions. Logic has nothing to do with it. The computer can work quite like the human brain and make equally flawed decisions, just much faster. It is just more consistant.

The problem is that you have a massively parallel analog device that you want to record. That device continues to modify its "applications" at will with new code. And the analog data cannot be quantified into set field lengths.

Example:

How do you store a smell? And then set a key to the smell that associates it with actions or then cross links it a whole other section of data that may not even be related. Just a fresh baked batch of chocolate cookies could be linked to various remembered and 'locked' memories at home. Then cross-linked to "grandma's". And the next time you smell the same, it may unlock a memory that really was linked, but the action changes. So, one time, you are walking down the street and smell cookies at the bakery. Depending on other environment variables, what memories it triggers, and current 'problem set' running, you may go into the bakery.

I would guess that even a century is not enough time to create anything that could map out the gathering tools or the decision ability in a computer to map a person.

Don't confuse computer with processor. Software would be he key factor here, because, in one way or another, i believe normal hardware is perfectly capable. We can still make what computers can (and woman tend to excel on that) witch is to multitask.

800 stream cores seem much ? I've catched my Ms, writing a sms, checking the mail in her lap warmer and watching tv all the same time. And she was relaxing. A computer would do it, with diferent dedicated modules.

Possible yes. Modular approach would be the key. Possible yes. The biggest problem is that such idea isn't profitable.

And thus, not doable.

 

[earthman]

The human brain is not a computer, its a biological organism that learns, adapts, sometimes rewires itself, and ultimately degrades in a way that is probably not possible to "simulate". Any attempt to record a person's thoughts would not yield a working simulation of a brain any more than a recording of someone's voice could carry on a conversation. Perhaps some of the older among you remember the Max Headroom episode where the crooked company was selling people time with a computer recreation of their dead loved ones, but it was really just video loops with generic responses. Perhaps the only way to duplicate a brain is to clone it, but that only duplicates the physical brain, not the complex experiences that make up a person.

 

[CycloWizard]

Originally posted by: gsellis
Actually, the computer only makes programmed decisions. Logic has nothing to do with it. The computer can work quite like the human brain and make equally flawed decisions, just much faster. It is just more consistant.

The problem is that you have a massively parallel analog device that you want to record. That device continues to modify its "applications" at will with new code. And the analog data cannot be quantified into set field lengths.

The brain isn't truly though - it uses digital signaling.

Example:

How do you store a smell? And then set a key to the smell that associates it with actions or then cross links it a whole other section of data that may not even be related. Just a fresh baked batch of chocolate cookies could be linked to various remembered and 'locked' memories at home. Then cross-linked to "grandma's". And the next time you smell the same, it may unlock a memory that really was linked, but the action changes. So, one time, you are walking down the street and smell cookies at the bakery. Depending on other environment variables, what memories it triggers, and current 'problem set' running, you may go into the bakery.

Smell is simply a series of chemical reactions, just like sight. The olfactory receptors have been pretty thoroughly mapped, though I don't know how easily they could be reproduced ex vivo. If you could quantify the olfactory inputs, as well as all other sensory inputs, then you would have all of the environmental inputs that a person would have. Then, the only thing left would be the person's mood, et cetera that are not necessarily related to external stimuli, though perhaps they could be mapped to hormones or something.

I would guess that even a century is not enough time to create anything that could map out the gathering tools or the decision ability in a computer to map a person.

I agree. Though it may be possible in principle, actually doing it is not so easy, to say the least.

 

[RESmonkey]

That's if we don't biologically attain immortality. I think messing with DNA/telomeres is the key to immortality. It is definitely possible to not age.

 

[aj654987]

Well they wouldnt know how well it worked until the first trials, I imagine it would take quite a few iterations for it to finally work right so thats a few generations after the technology started before it gets better.

For example, the very first time they try it, they are still learning how to do it and dont know what data is needed and what isnt. Imagine they wait 20 years for the first test patient to die only to realize they didnt record some key piece of data that they didnt realize was needed.