Author Topic: A Better Human  (Read 8434 times)

Offline BearMills

  • Hirundo Rustica
  • *****
  • Posts: 1135
  • Karma: +30/-2
  • Grizzly
    • View Profile
Re: A Better Human
« Reply #40 on: June 10, 2008, 07:42:00 PM »
I LIKE STAYING OLD SCHOOl
Co-Founder

Offline Milquetoast

  • Patas
  • ***
  • Posts: 178
  • Karma: +16/-3
  • ... was Milquetoast but not anymore...
    • View Profile
Re: A Better Human
« Reply #41 on: June 10, 2008, 10:10:48 PM »
lol. this thread is gettin a little complicated. and im a computer major  ;).

anyways i believe it was the mayans who predicted that the world as we know it will end in December of 2012, i believe December 21st but im not sure, and a new race of sentient beings will take over the planet. What if its the robots. What if someone makes a huge breakthrough in AI, some makes a working self progressing artificial intelligence. the next race of sentient beings could be robots.

omg  :o :o :o

lol. i don't really believe this btw. though i do believe in the possibility of AI being able to run without human interaction is approaching. all it really takes is one major breakthrough in software design, like a new algorithm that no ones thought of yet.

Well, I have been in the software engineering field now for about a decade and a half now (if you include the years in college) and I think it's going to take a little more than a single major breakthrough in either design methodology or algorithms.  I mean, we're talking about simulating a human brain here.  Not building a faster searching algorithm or even proving P = NP.  Besides, there's also a problem of loading the data / emulating the human experience in order to even test a system.  For example, Neural Networks are pretty good AI models that show a lot of promise.  However, there are countless learning paradigms associated to train machines because no one's quite sure about the best way to teach machines.

Although, despite my doubts, I have to say that Kurzweil deserves some credit for putting his money where this mouth is:

http://www.longbets.org/1

For those who don't know, the Turing Test is the test (proposed by the Father of Computer Science, Alan Turing) to see if an AI has reached the level of human intelligence.

Offline Andrew Robillard

  • Mangabey
  • ****
  • Posts: 259
  • Karma: +17/-11
    • View Profile
Re: A Better Human
« Reply #42 on: June 11, 2008, 07:13:12 AM »
lol. this thread is gettin a little complicated. and im a computer major  ;).

anyways i believe it was the mayans who predicted that the world as we know it will end in December of 2012, i believe December 21st but im not sure, and a new race of sentient beings will take over the planet. What if its the robots. What if someone makes a huge breakthrough in AI, some makes a working self progressing artificial intelligence. the next race of sentient beings could be robots.

omg  :o :o :o

lol. i don't really believe this btw. though i do believe in the possibility of AI being able to run without human interaction is approaching. all it really takes is one major breakthrough in software design, like a new algorithm that no ones thought of yet.

Well, I have been in the software engineering field now for about a decade and a half now (if you include the years in college) and I think it's going to take a little more than a single major breakthrough in either design methodology or algorithms.  I mean, we're talking about simulating a human brain here.  Not building a faster searching algorithm or even proving P = NP.  Besides, there's also a problem of loading the data / emulating the human experience in order to even test a system.  For example, Neural Networks are pretty good AI models that show a lot of promise.  However, there are countless learning paradigms associated to train machines because no one's quite sure about the best way to teach machines.

Although, despite my doubts, I have to say that Kurzweil deserves some credit for putting his money where this mouth is:

http://www.longbets.org/1

For those who don't know, the Turing Test is the test (proposed by the Father of Computer Science, Alan Turing) to see if an AI has reached the level of human intelligence.

thanks for your post. it is very interesting and i will post a more detailed response as soon as i finish reading it.
btw, what kind of software engineering do you do, like what have been your past job requirements, fixxing software for a company, designing new software, working on AI, building new algorithms, making OS, etc?

EDIT: ok. heres my thoughts. personally i side with Kapor. I think it can be done.

"The brain is self-organizing, which means that it is created with relatively little innate knowledge. Most of its complexity comes from its own interaction with a complex world. Thus it will be necessary to provide an artificial intelligence with an education just as we do with a natural intelligence."

because of this i think that if someone creates an algorithm that learns based on its environment and has a basic understanding of emotions that can develop based on its observations of other humans displaying emotion. If someone could create this algorithm you could create "life experiences" for the AI that it would "learn" from and develop into a more intelligent being.
The thing is that according to Kapor the AI would develop its intelligence at a much faster rate than humans can and its intelligence would continue to grow exponentially.

also the article mentions that there have been AI's that have passed closed versions of the turing test, i.e. tests confined to a specific topic. This means that there are already self sufficient AI's that can do specific jobs, just no AI's that can emulate the full capacity of a human and the human brain.

at least thats what i got from the article.

EDIT: I agree with Kurzweil not Kapor. sorry. got the names mixed up.
« Last Edit: June 11, 2008, 09:37:36 PM by Robillard »
"With the will and the passion impossible is nothing."

"I've missed more than 9,000 shots in my career. I've lost almost 300 games. 26 times, I've been trusted to take the game-winning shot and missed. I've failed over and over and over again in my life. And that is why I succeed."
-Michael Jord

Offline Milquetoast

  • Patas
  • ***
  • Posts: 178
  • Karma: +16/-3
  • ... was Milquetoast but not anymore...
    • View Profile
Re: A Better Human
« Reply #43 on: June 11, 2008, 08:47:29 PM »
thanks for your post. it is very interesting and i will post a more detailed response as soon as i finish reading it.
btw, what kind of software engineering do you do, like what have been your past job requirements, fixxing software for a company, designing new software, working on AI, building new algorithms, making OS, etc?

EDIT: ok. heres my thoughts. personally i side with Kapor. I think it can be done.

"The brain is self-organizing, which means that it is created with relatively little innate knowledge. Most of its complexity comes from its own interaction with a complex world. Thus it will be necessary to provide an artificial intelligence with an education just as we do with a natural intelligence."

because of this i think that if someone creates an algorithm that learns based on its environment and has a basic understanding of emotions that can develop based on its observations of other humans displaying emotion. If someone could create this algorithm you could create "life experiences" for the AI that it would "learn" from and develop into a more intelligent being.
The thing is that according to Kapor the AI would develop its intelligence at a much faster rate than humans can and its intelligence would continue to grow exponentially.

also the article mentions that there have been AI's that have passed closed versions of the turing test, i.e. tests confined to a specific topic. This means that there are already self sufficient AI's that can do specific jobs, just no AI's that can emulate the full capacity of a human and the human brain.

at least thats what i got from the article.

Unfortunately, I haven't worked with anything nearly as exciting.  Outside of academia, there really aren't that many R&D jobs going around.  All of my jobs in the past have been designing, implementing and maintaining large enterprise applications to spec.  However, I do still remember much from my AI, OS and Algorithms classes and have been trying to keeping abreast of new developments in the fields.  Now, before I get carried away on another long post, I just want to say up front that my experience by no means qualifies me to make a better guesstimate than anyone out there.  I'm just making a random guess like everyone else, and there is a good chance that I could be very wrong.

I think you mean you agree with Kurzweil, right?  Kapor predicted that no computer will have passed the Turning Test by 2029.  You and Kurzweil may very well be right...  20 years is a long time and there could be amazing developments in this period.  In fact, think of it this way: if, in 5 years' time, we could come up with enough hardware to simulate a blank human brain and enough understanding of the human brain to come up with an exact replica of how human brains learn, then we would have a whole 15 years to "raise" this computer AI.  Even if all we could manage to do is raise it at normal human pace, we'd still have a 15-year-old "brain" by 2029.

However, as of right now, my money is still on Kapor.  Again, I do believe such a development is an inevitability, but 20 years feels a little too aggressive to me because there is still so little we know about how the human brain works.  But don't get me wrong -- I hope Kurzweil is right.  I hope he is right because I'd much rather live in an exciting, fast-paced world than the current one I'm living in.  :)

Offline Andrew Robillard

  • Mangabey
  • ****
  • Posts: 259
  • Karma: +17/-11
    • View Profile
Re: A Better Human
« Reply #44 on: June 11, 2008, 09:58:53 PM »
ya. i really want to do R&D but im probably not going to become a professor, don't like teaching and i kinda want to get paid better  :P, so we'll see what happens. The article mentioned something about humans having fully virtual experiences in the near future which is something i've wanted to work on since i was a kid. AI, nanotechnology, and biotechnology i think it would be called (like cyborg sort of stuff, i.e. the blind lady that just recently got an implant in her brain connected to a camera that has so far allowed her to see a few dots and patterns of light) have also interested me a lot too.

but we'll see where i wind up by the end of college.

something interesting to see with the AI though would be to explore whether or not different AI's with the same exact algorithm but possibly different learning environments would develop different "personality traits". It'll be interesting to see to what extent an AI will be able to mimic humanity or if it will just become one central AI with a bunch of "drones" as a result of a network.

also i wonder if humans or AI ever become directly linked to each other through some sort of internet type network whether or not a separate consciousness would develop as a result of so many consciousness' communicating with each other continuously. Even if its something thats barely there that we aren't really aware of but has some sort of impact on thought process.
"With the will and the passion impossible is nothing."

"I've missed more than 9,000 shots in my career. I've lost almost 300 games. 26 times, I've been trusted to take the game-winning shot and missed. I've failed over and over and over again in my life. And that is why I succeed."
-Michael Jord

Offline BearMills

  • Hirundo Rustica
  • *****
  • Posts: 1135
  • Karma: +30/-2
  • Grizzly
    • View Profile
Re: A Better Human
« Reply #45 on: June 12, 2008, 06:19:36 AM »
ya. i really want to do R&D but im probably not going to become a professor, don't like teaching and i kinda want to get paid better  :P, so we'll see what happens. The article mentioned something about humans having fully virtual experiences in the near future which is something i've wanted to work on since i was a kid. AI, nanotechnology, and biotechnology i think it would be called (like cyborg sort of stuff, i.e. the blind lady that just recently got an implant in her brain connected to a camera that has so far allowed her to see a few dots and patterns of light) have also interested me a lot too.

but we'll see where i wind up by the end of college.

something interesting to see with the AI though would be to explore whether or not different AI's with the same exact algorithm but possibly different learning environments would develop different "personality traits". It'll be interesting to see to what extent an AI will be able to mimic humanity or if it will just become one central AI with a bunch of "drones" as a result of a network.

also i wonder if humans or AI ever become directly linked to each other through some sort of internet type network whether or not a separate consciousness would develop as a result of so many consciousness' communicating with each other continuously. Even if its something thats barely there that we aren't really aware of but has some sort of impact on thought process.

My future~
WWF Photographer (World Wildllife foundation)
Zoologist-studying Monkeys and Cats
MarineBiology- Dont know yet
Co-Founder

Offline Milquetoast

  • Patas
  • ***
  • Posts: 178
  • Karma: +16/-3
  • ... was Milquetoast but not anymore...
    • View Profile
Re: A Better Human
« Reply #46 on: June 14, 2008, 02:08:19 AM »
This is going to be a long one, Robillard, so...  My apologies in advance.  Yeah, the future does sound exciting, does it not?  Although, I wonder if kids in the future will find all this stuff to be no different from how we see a TV or a personal computer (actually, I'm sure they will) and keep looking forward to the next big thing.  That is human nature, is it not?

I'm sure that eventually AIs "born" with the same algorithm but "brought up" in a different environment will start exhibiting differences.  However, how much difference they'll exhibit and how different the environment will have to be will depend on how many sensory inputs they have and whether or not they "learn" from all of them.  Having said that, however, my guess is that "snapshots" will be taken of fully-functional, mostly-educated AIs and sold / distributed like any other OS release.  These AIs will have identical personalities out of the box, but their personalities will slowly drift apart as they learn from their operating environment (basically, the Users)...  Until they get so bloated with virii, spyware and broken registry keys that the User gets fed up and does a wipe-and-reinstall.  ;)

As for the drone vs. individuality question, it will depend on technological advancement in the field.  We already have drone AIs with limited intelligence out there.  That's how Google and every other major search engine builds their indexes.  They send out webcrawlers that scour the web for pages and build this huge full-text index based on what the crawlers find.  Based on certain rules, the "hivemind" determines how relevant a particular page is with a particular keyword.  We'll have to wait a bit longer for functional AIs with individual (and evolving) personalities, I'm afraid...

Technically, humans are already directly linked to each other via the Internet.  It's just that the interface is still very primitive and clunky.  There are also, arguably, already countless separate "consciousnesses" developed as a result of the connections.  Take APK, for example.  By virtue of the Karma system, Mod system and -- most importantly -- good old fashioned praise, shame and ridicule, a tone is constantly being set.  Ideas and various forms of behavior are constantly being reinforced or shot down.  If one were able to read every post, every Applaud / Smite and every Moderation, they'd be able to understand the "consciousness" of this system and be able to come up with a plan to game the system.  Even though you and I do not have access to detailed stats and do not read every single post (at least, I don't and I hope you don't either  ;)), we learn enough about what is acceptable and what is not on this board and act accordingly.  In other words, the feedback from APK already has an influence on our thought processes.

But, I think I know what you're getting at.  Will humans ever be directly jacked in?  I think it's only a matter of time before we're technologically capable, but whether or not we'll get through the ethical implications and whether or not it will be desirable is yet to be seen.

Offline Clive Vella

  • Oryctolagus Cuniculus
  • *
  • Posts: 26
  • Karma: +2/-0
    • View Profile
Re: A Better Human
« Reply #47 on: June 14, 2008, 05:50:31 AM »
I haven't had the chance to read the full thread, but have you guys mentioned the mental part of being 'naturally attractive/better' to procreate a new survivable generation?

A man who is strong and agile, and has all the characteristics of what modern times consider a good-looking man doesn't mean he would be the first choice as a parent to a new generation of men.

A man who is all of the above, and also has a strong character and will, will most probably always be the first choice by anyone to follow or to procreate with. (If any of you ever heard about being an "Alpha Male", that is what I'm trying to point out)

So even though beauty will become artificial (I think someone mentioned plastic surgery) there will still be a major factor affecting who will be chosen or not, as mankind is also a very mental and emotional being, and in a society like ours it is considered very important, even though many neglect a proper way-of-mind.

Felt to urge to point these out, even though it was quite a mediocre explanation, also my first post. Hope I did not run too off the topic, but I am very keen on the subject even though I have no academic background.




Offline Terry McIntosh

  • Patas
  • ***
  • Posts: 220
  • Karma: +19/-7
  • Tear It Up Terry
    • View Profile
Re: A Better Human
« Reply #48 on: June 14, 2008, 09:32:51 AM »
o.0
<.<
>.>
 ???  [WTF]  ???

all this talk of humanity evolving or having robots take over... i'll just cross that bridge if a have to because im sure i will be dead when it becomes an issue.
I like to touch different kinds of surface textures.
Parkour lets me do that without feeling weird.

Offline Andrew Robillard

  • Mangabey
  • ****
  • Posts: 259
  • Karma: +17/-11
    • View Profile
Re: A Better Human
« Reply #49 on: June 23, 2008, 08:24:51 PM »
this is a very interesting video i found, eugene and anyone else interested may want to take the time to sit down and watch the thing all the way through (yes i know its long but well worth it) maybe watch it in segments cause my head hurts now lol.
anyways he offers interesting data and ideas about the exponential growth of technology and evolution, including a lot of research and data to back up his claims. he talks about the paradigms and overcoming them and patterns inherent in overcoming them.
and about the human brain being complex and having to figure it out for AI, he addresses that too. mostly along the lines of once we understand one section, the next becomes that much easier to map, which makes the next one even easier, etc. growing exponentially till we have the entire brain mapped out.
he also predicts that we will probably have chips (micro computers) implanted in our retina, clothing, skin everything creating fully virtual worlds were we can see virtual objects and interact with them. change our surroundings and fully immerse our senses creating the feeling we're actually there and be able to convey though emotions to other people, be able to just plug in and feel the emotions of someone else, all by the 2020's.
He also describes technology we have today that has cured type 1 diabetes in rats. basically synthesized red blood cells that control the release and flow of insulin in the body. this is also supposed to be able to improve the efficiency of our body and allow us to do things like stay underwater for hours at a time without a taking a breath.
it's very interesting stuff, sorry i don't know who the speaker is but he is speaking at the TED convention, i think its called, held in Canada like every year i think.


<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/IfbOyw3CT6A&hl=en"></param><embed src="http://www.youtube.com/v/IfbOyw3CT6A&hl=en" type="application/x-shockwave-flash" width="425" height="344"></embed></object>

enjoy.

EDIT: and to add one more 20 min video. here is one about brain theory and intelligence. Basically that intelligence isn't based on behavior (like the turing test suggests) but rather that intelligence is based on predictions that we make from our memories of past events. its another really good watch, the speaker is Jeff Hawkins.
<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/G6CVj5IQkzk&hl=en"></param><embed src="http://www.youtube.com/v/G6CVj5IQkzk&hl=en" type="application/x-shockwave-flash" width="425" height="344"></embed></object>



and finally, i think. a much shorter video. only about 6 minutes, about machines that actually evolve and become self aware, kinda like i suggested earlier, rather than being programmed and already having a plethora of knowledge.
<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/lMkHYE9-R0A&hl=en"></param><embed src="http://www.youtube.com/v/lMkHYE9-R0A&hl=en" type="application/x-shockwave-flash" width="425" height="344"></embed></object>
« Last Edit: June 23, 2008, 09:17:51 PM by Robillard »
"With the will and the passion impossible is nothing."

"I've missed more than 9,000 shots in my career. I've lost almost 300 games. 26 times, I've been trusted to take the game-winning shot and missed. I've failed over and over and over again in my life. And that is why I succeed."
-Michael Jord

Offline Milquetoast

  • Patas
  • ***
  • Posts: 178
  • Karma: +16/-3
  • ... was Milquetoast but not anymore...
    • View Profile
Re: A Better Human
« Reply #50 on: June 30, 2008, 03:39:28 PM »
Thanks for the links to the videos, Robillard.  TED is an excellent (and I would argue extremely important) annual conference in California.  The videos are very enlightening, and while I'm still not convinced of Kurzweil's projected timelines, I feel even more reassured of the inevitability of it all.  Well, at least we're just 2 years away from his closest predicted milestone.  In the video, he predicted that we'll get computers stitched into our clothing by 2010.  I've read about prototypes of this for a while now, so I'll assume he meant commercially.  We'll know if he's accurate or not in about 2 years.

I think Hawkins might be onto something.  Our "old view" of intelligence as behavior might be exactly what's throwing off research into intelligence.  However, even if he is correct, I don't believe that invalidates the Turing Test in any way.  The reason I say this is because the Turing Test is not meant to test if a machine actually behaves and reacts identically to a human being, but rather if a machine can pretend to do so.  In other words, it's kind of like a game where the machine is trying to outsmart and fool the judges into thinking that it's a human being.  In fact, in order for a machine to pass the Turing Test, it will have to do exactly what Hawkins suggests we do -- predict what a human would do and act accordingly.