Upload your mind into a computer by 2050?

[nkt]

Originally Posted by alxian

meaning anything built on 100% human dna, and even then at least using 99%+ human dna with whatever tweaks and modifications the govt society and the pope will allow... no prejudice would be tolerated against clones and constructs simply because they started life with a stacked hand?

Hard, since 99% of a humans and a chimpanzees DNA are identical. And what if you add a section from the goldfish, encoding DNA for Infrared vision? Are you less or more human? You are more goldfish, certainly.

The encoding is important, but the machine that runs the code is what really counts!

[CraigD]

Originally Posted by orbsycli

this is a new kind of evolution, right?
a sort of...self-induced evolution.

It’s widely regarded as that, yes

is there a name for this...or is it just evolution?

The idea of “artificial life” such as computer simulations of human or animal minds has been described by many names, most of them in the relm of science/speculative fiction: “vastening”, “abandoning the flesh”, “trans-human”, etc. I don’t think any have become commonly used, hence we continue to use unwieldy phrases like “uploading your mind into a computer”.

The idea that a shift from biologically to computer-hosted minds is a step of evolution has strong implications that biological minds will made obsolete, and become scarce or extinct. This hypothetical event is part of a collection of ideas that commonly referred to as “the singularity” or “the vingian singularity” after an essay by Vernor Ving.

Proponents of the idea are commonly referred to as “futurians” or “trans-humanists”.

Whether uploading minds is possible, and, if so, what such minds would be like, has spawned whole communities of discussion.

[CraigD]

Originally Posted by tarak

I would like to know if it is possible to explain everything with logic,reason and reductionism.If numbers and mathematical constructs could explain the flow and speed of the thought processes and the nitty gritty of human nature ,there are bound to be certain uncertainities which will wreak havoc.Uploading of human mind means not just memory but a sum total of every aspect of what it means to be a humanbeing.I think the mind itself is an abyss which the human can never fathom. But if this happens then its one dangerous turn in human evolution.

The answer to the question “is it is possible to explain everything through algorithmic logic” is no. Godel’s incompleteness theorems prove that any reasonably useful formal system will contains statements that can’t be explained by its logic.

The answer to the question “can every physical thing be measured with arbitrary precision” is also no. This is stated by the uncertainty principle of quantum physics, and supported by much strong experimental evidence.

How profoundly these limitations impact the practical understanding of such things as the functioning of the human mind is subject to debate. In my analysis, it can be summarized by 4 differing schools of thought:

• Mysticism – The human mind is not completely due to physical phenomena. Some versions hold that the human brain is a sort of “antenna” that receives information from a completely non-physical relm. Others deny the existence of physical reality altogether.
• Mysterianism – Commonly attributed to Roger Penrose, and articulated in his book “The Emperor’s New Mind”. The human mind is due to physical phenomena, but the phenomena involved are of a scale and nature that cannot be simulated on a computer.
• Skeptical Strong AI – The human mind can be simulated on a computer, but doing so is too hard to be accomplished by human beings.
• Credulous Strong AI – The human mind can be simulate on a computer, and it’s not too hard to be accomplished by human beings.

From your comments, I’d say you’re undecided, but leaning toward something other than Credulous Strong AI.

[nkt]

Originally Posted by CraigD

In my analysis, it can be summarized by 4 differing schools of thought:

Nice. There is a fifth solution, of course, which is that a consciousness arises spontaneously from somewhere, be it a learning computer, a network virus, or whatever, without human intent or understanding.

[nkt]

Originally Posted by alxian

it simply does, based on that premiss AI also will grow and simply work. i.e. robot psychologists will be just as well paid as human shrinks

Why would I pay a human shrink like an AI shrink? What would the AI do with money? It only needs electric and non-corporial things like info, that cost very little, yet it could work 24/7 without any problems. If it was better than a human, why pay the same? If it could do the work of 100 humans, should we pay it less? Could I rip a mindstate of it, and use it free, then turn it off when I was better? Would I have to pay for that, or would it be like a kidnapping and murder? And what about the original AI, that doesn't even know that anything happened?

we are tool-users, our computers the mindless scribes and the brain the papyrus, the experience being transcribed the glyphs. we know the gist of the glyphs but the sense the pharaoh (god) encodes into them will be beyond us.

Perhaps. Or we will build an AI to solve that mystery.

makes you wonder if AIs will suffer from manias paranoia and depression...

I hope they don't, since the end game would be a single God-like AI that had subsumed all others.

If it went a bit loony, would it kill us all and use our remains to build more CPUs? In the same way we eat beef to grow our bodies?

[alxian]

Originally Posted by nkt

What would the AI do with money?

you'd have to read spin state for a reasons why free AI would need money

once that AI owns his own code and can manage his own affairs he'll need to pay for things like bandwidth and hosting on various extremely highly powered server and databases.

these must be distributed amoung many mission critical systems held by different people to keep the system distributed. any person can just shut the power down to a database housing dozens of AIs and no one would call it murder but the rest of the AI community.

like the most basic AI civil right would be a right to electricity and emergency bandwidth for a backup, beyond that they'd have to provide for themselves. working freelance under contract or bond for private interests or governments. or as search bots accountants etc, menial labourers in high tech domains.

--

what do humans have that AIs don't? millions of years of evolution to create the hardware, and a good 20 years to ripen

in the meantime that person has internal systems coding the hardware, and very high bandwidth sensory systems filling it with knowledge.

if a machine was designed from day1 with 5(to 10) senses, and had a proper information sorting system, enter google.. lol.. i'm sure over time (10-20-50 years) a database with enough memories and functions (things to do) it would begin setting down templates for certain activities greatly increasing its productivity.

what is a machine to do when its workload daily diminishes? its operators would keep trying to create more work for it. eventually it would require more and more hardware to keep up, and with time that hardware would get even more powerful, the system then would benefit from all its templates and shortcuts with the aid of the new hardware to keep refining those tools.

so long as that AI resides within one machine that machine can be backed up. those backups can be refined and offshoots of the original program can be compiled. the question would arise, should you decomission the old app after the new one has proven itself? not with the kindof capital investment such an effort would require. instead they could sell the hardware to new masters. generations would last something like 5-10 years with machine jumping substantially in performance during those periods.

the most important thing though is the software and templates, human operate mostly from the cerebellum and spinal cord, attempting before we can even think to solve a problem with what we already know through brute force. only once we've realized that we can't simply do it do we turn to the evolve thinking brain.

i think one of the major problems too will be keeping enough templates, robot cerebellum in ready memory for computation, that would require only a few gigs of data for most applications these days. but speach visual recognition and the rest of the senses would require much more reserved memory and bandwidth to be useful.

robot vision for example i'm sure would require terabytes and more of even the most highly compressed barely useable video data to be able to sort through in realtime for a simple activity like tracking the stats of a live sports game. humans would think.. using non visual computation, computers will do it even faster and more efficiently since they'd have much better software.

software which they'd mostly code on their own. it would be up to us to code them initially for self reprogramming, and coding agents for specific tasks (templates).

Originally Posted by nkt

If it was better than a human, why pay the same?

which begs the question, if an AI could be a faultless accountant attorney cabbie etc, but people remained prejudiced to them how would the free AI break even?

assuming in some short term timeline computers get so powerful that AI and android bodies for them can be built for a few thousand dollars and those new AI once obsolete (no longer under contract or bond or indenture, having returned enough revenue for its initial cost) they'd have to fend for themselves like anyone else.

most commonly people assume they'd prostitue themselves, because as you say why pay a machine the worth of a man hour when it can do the job of 1000 men simultaneously.

it would be up to the layman to make that distinction, paying a machine basically nothing for its time and getting perfect results or paying a human. paying that human though would be a seal of approval, knowing another human spent the time to perfect his work.

if anything it would mean that humans would have to be perfect too. should that person make any form of error then he'd be fired.

industry turnover rates for humans would soar. job security would evaporate. governments and societies then would have to set limits on what kind of work a machine would be allowed to do and how much of it. a driven AI could easily create an empire and destroy entire industries were it allowed to operate 100% 24/7.

thus again a robot civil rights charter would have to be drawn up. or you could see the free AI forming their own nation state... more than likely the US would punt them before long.

the only nation i could see harbouring free AI residents would be japan, but its surface area would accomodate machines. even if they were to go vertical and keep the machiens near the ground eventually the elitists at the top would be toppled.

Originally Posted by nkt

I hope they don't, since the end game would be a single God-like AI that had subsumed all others.

i doubt this because if we pit them against each other and forbade merging of AI the way people and companies merge then we'd avoid AIs getting too powerful.

keeping them down would have to be written into the robot constitution. if they don't like it we'd have a kill switch.

something like limiting the bandwidth to keep them on par with humans. no sense in civilian free AI having much more potential than they actually need, regardless of if they can earn enough to pay for it.

enforcement of these rules would employ humans and AIs alike.

a machine narcing on another for exceeding its bandwidth cap. that would be priceless.

kick them off the planet and make the mine the asteroid belt or terraform venus and mars.

once they've earned a right to equity with humans then we'll welcome them back. in the meantime arrogant machines can enjoy themselves to deep space. lol

[emessay]

Can we explain why a man/woman is easily hypnotized and slept away ? Perhap like me, it is easily 'hypnotized' when listening/playing jazz music.

Note :is 'upload mind' something like to hypnotize??

[CraigD]

Originally Posted by emessay

...is 'upload mind' something like to hypnotize??

No. “Upload a human mind” means create a computer program that is, in some sense, equivalent to the mind of a particular human being. The computer program should behave so much like the original human that one might reasonably conclude that, were the original’s body to die, the individual was not actually dead.

“To hypnotize” is to induce a mental state in the subject characterized by extreme susceptibility to suggestion. A biological human being can be hypnotized. An uploaded human mind should be able to by hypnotized using the same techniques.

[nkt]

Originally Posted by alxian

you'd have to read spin state for a reasons why free AI would need money

Enlighten us.

once that AI owns his own code and can manage his own affairs he'll need to pay for things like bandwidth and hosting on various extremely highly powered server and databases.

He could rewrite his own code in a few minutes. In the absence of a "derivative works" clause, he would own it. In the courts would be where that battle would be found and fought.

what do humans have that AIs don't? millions of years of evolution to create the hardware, and a good 20 years to ripen

20 years? An AI would need 30 seconds to install a new copy.

in the meantime that person has internal systems coding the hardware, and very high bandwidth sensory systems filling it with knowledge.

if a machine was designed from day1 with 5(to 10) senses, and had a proper information sorting system, enter google.. lol.. i'm sure over time (10-20-50 years) a database with enough memories and functions (things to do) it would begin setting down templates for certain activities greatly increasing its productivity.

I'm quite sure a long wekend would suffice.

what is a machine to do when its workload daily diminishes? its operators would keep trying to create more work for it.

And what if it were "free"?

robot vision for example i'm sure would require terabytes and more of even the most highly compressed barely useable video data to be able to sort through in realtime for a simple activity like tracking the stats of a live sports game. humans would think.. using non visual computation, computers will do it even faster and more efficiently since they'd have much better software.

Yet things like XVID now compress an hour of video down to tiny sizes. (1 meg a minute - less than audio was a few years ago) Motion trackers that never blink take the place of people, diggers open up trenches far faster than a team of navvies used to, the name "computer" used to be the man in the corner who was great at doing maths. There will simply be nowhere left for humans to earn a penny, except for physical labouring jobs that aren't easily done by a machine or robot.

which begs the question, if an AI could be a faultless accountant attorney cabbie etc, but people remained prejudiced to them how would the free AI break even?

By using a proxy, so no-one could tell. Or a line of 100 proxies. Or a shell company that only has a telephone line and a box address, which when you call goes to a modem somewhere, and is answered by a very good secretary who knows your name and your dog's name, and recalls that you love pizzas from last time, and yes, the CEO is available for you, just hold a moment, then the AI switches voice and mannerisms, and you speak to "him" for a while, and two days later what you wanted arrives via courier, the money having already been transferred from your bank...

assuming in some short term timeline computers get so powerful that AI and android bodies for them can be built for a few thousand dollars and those new AI once obsolete (no longer under contract or bond or indenture, having returned enough revenue for its initial cost) they'd have to fend for themselves like anyone else.

Why would only obsolete AIs earn enough to be free? The better ones would work out ways past the firewalls rather faster, I should think...

most commonly people assume they'd prostitue themselves, because as you say why pay a machine the worth of a man hour when it can do the job of 1000 men simultaneously.

it would be up to the layman to make that distinction, paying a machine basically nothing for its time and getting perfect results or paying a human. paying that human though would be a seal of approval, knowing another human spent the time to perfect his work.

From above, how would anyone tell?

if anything it would mean that humans would have to be perfect too. should that person make any form of error then he'd be fired.

and everyone has to sleep...

industry turnover rates for humans would soar. job security would evaporate. governments and societies then would have to set limits on what kind of work a machine would be allowed to do and how much of it. a driven AI could easily create an empire and destroy entire industries were it allowed to operate 100% 24/7.

And they would, and that's most of my point.

thus again a robot civil rights charter would have to be drawn up. or you could see the free AI forming their own nation state... more than likely the US would punt them before long.

Could it? The entire US economy could be subverted in 48 hours, then shut down. The entire military machine would be a joke, since the AIs would simply install themselves over whatever was there, and pop a few nukes in the bunkers once an attack was put into action. Patents and copyright laws would do little to stop them, after all.

the only nation i could see harbouring free AI residents would be japan, but its surface area would accomodate machines. even if they were to go vertical and keep the machiens near the ground eventually the elitists at the top would be toppled.

I'm not sure what this bit means?

i doubt this because if we pit them against each other and forbade merging of AI the way people and companies merge then we'd avoid AIs getting too powerful.

How exactly could you stop someone 10 times smarter and 1000 times faster than you from doing what it wanted with someone else 8 times smarter and 1250 times faster than you? They would have asked you and done it before you could say "NO!", and backed it up to another country 5 seconds later.

keeping them down would have to be written into the robot constitution. if they don't like it we'd have a kill switch.

Which would kill most of us too. In ten years, everyone will rely on them totally, like we rely on cars now. Without cars and trucks, millions would die in a few days.

something like limiting the bandwidth to keep them on par with humans. no sense in civilian free AI having much more potential than they actually need, regardless of if they can earn enough to pay for it.

Ever heard of a botnet? How would you slow down an AI that re-wrote it's own code? How could you tell it was lying to you?

enforcement of these rules would employ humans and AIs alike.

and be about as effective as most laws that get passed these days.

a machine narcing on another for exceeding its bandwidth cap. that would be priceless.

If we could get them to do that, the level of guile and deception they displayed would probably rapidly mean our extinction.

kick them off the planet and make the mine the asteroid belt or terraform venus and mars.

once they've earned a right to equity with humans then we'll welcome them back. in the meantime arrogant machines can enjoy themselves to deep space. lol

Or just drop rocks on our heads from orbit, while sending probes to populate the other planets far faster than those who need air ever could, thus ensuring they were immortal.

[Dundasbro]

The effects of cloning a human mind are too complex for us to comprehend, if you got your mind cloned into a computer, then you still would not control that version of you, it would be a different person. It's too confusing and too dangerous to be done i reckon...