Discussing the "bigger picture?"


#1

just started mapping yesterday, already loving THAT the way it combines science, art, gaming and more... im thinking about a thread to chat abouot the larger potential to Humanity should it succeed as intended... see ya on the leaderboard wink


#2

Well the most popular big picture ideas generally have to do with curing neural disorders.  Being a computer science guy, I tend to prefer the AI perspective.  Can we figure out what makes the brain tick?  Can we reproduce it?  The applications of AI are pretty limitless.


#3

Exactly where I'd hoped to start, Matt... if the neurological connections prove to be the "archive of ME" then the potential for AI immediately expands. One aspect could be the possible preservation of personality (a la "Dollhouse") and another could be focusing AI development efforts toward real cognition!


#4

One way that I look at it is this; we can only truthfully judge the universe based on what we can observe, therefore if we can observe nothing more than the components of the brain, it follows that the phenomenon of the mind must exist as a function of that structure.


#5

I would say that I try to be reserved in my expectations for the future.  I think Dollhouse and "real cognition" are still a long ways off.  But I do think that with a little bit of a push, computers can be improved by working in a more brainlike fashion.  For example, we've hit a barrier in terms of making chips faster, and we are approaching some hard barriers in terms of making chips smaller.  The only thing that's left is using more chips together at the same time.  The problem with that approach is that the more stuff you have, the further away from each other it all is.  Very quickly, you lose all the gains you made by adding stuff.  This is roughly why your 4 core processor doesn't make programs 4x faster than a 1 core chip.


If there's one thing that the brain is good at, it's using lots of processing power in parallel.  I'm sure there's a lot of interesting stuff to be learned about how the brain economizes the transfer of information from one region to another distant region.

There are other things that humans do much better than machines which could revolutionize computing if we could replicate them.  The most interesting one for me is remembering.  You may sometimes feel like you are bad at it, but what you are doing is far more complex than pulling up data off a hard drive.  You are selecting very specific information based on very abstract & complex relationships.  You can even use these abstract relationships to synthesize new knowledge/information if you are cleaver enough.  This is sort of the foundation of learning, and if we could make a learning machine...

#6

http://www.huffingtonpost.com/2012/06/26/rats-study-animals-laugh-tickled-video_n_1627632.html?ncid=webmail2 validates our work here... rodent brains may be much more like ours than previously thought!


#7

Brain-like AI should be absolute priority, high above anything else, IMO. Everything we do begins as a set of taught routines. Our goal should be to enable smaller or bigger AI modules to perform these same routines in concert. After that it becomes a matter of scale. Everything we currently do could be done with much less effort thus transforming the world unprecedentedly, freeing us and making us sane and more civilized.


#8

Yes, as long as Asimov's safety protocols are in place! These routines combined with the simplest roomba or cockroach-type robots would be a huge step.


#9

But remember that Asimov's rules were not a suggestion of how we should do things, but a cautionary tale that even the best laid plans...


#10

Well-said, balkamm... the show SciFi Science has a scary episode about AI Uprising...


#11

What happens in the episode?


#12

Basically, they're gonna win! No matter what failsafes we employ, they'll outsmart them (a la Matrix...)


#13

I think people make a lot of unwarranted assumptions when trying to predict the future of robot uprisings.


1) We are capable of making robots that are capable of bringing about our destruction
2) There aren't any hard limits on "intelligence."  We know that there are hard limits on computability, why should intelligence/creativity be any different?
3) That robots would be belligerent.  In order to do all the things that we don't want to do, it's not necessarily the case that we would need to make machines intelligent enough to care that they are doing menial work.
4) That NLP == intelligence.  Just because a program can speak or understand speech does not necessarily mean that it's intelligent.  Siri and Google Now are getting much better in terms of understanding speech and intent.  I'm not afraid that they are going to kill me (or even refuse to do what I ask)

#14

Hi, Matt, thanks for your input. #s 1 and 2 are given, I agree absolutely. #3 seems to be the intriguing factor (again, the AniMatrix has an excellent "history" of the uprising)... the entire unrest is a result of discontent. If discontent is not a possibility, then neither is conflict! A wise teacher once touted "ecstasy is a result of welcoming discontent" which I took to mean that we have to PURSUE happiness rather than settle for good enough. If you deny both ecstasy and discontent, then menial work would remain possible. This incorporates #4 in that intelligence and AWARENESS are separate levels... which comes back to refute #2? Must we try to limit intelligence to prevent discontent? Would that even be possible if they are building each other?


#15

Recently watched "Paycheck" based on a Philip K. Dick story... eliminating memories by shooting synapses like a video game!


#16

On earlier thread: Dr. Michio Kaku insists the intelligence will inevitably surpass our own


#17
backupelk : " One way that I look at it is this; we can only truthfully judge the universe..." 
We loose when we believe in any truthfully judgement. Nobody can have a truthfully judgement about things in this Universe, due Godel's theorem: Nobody can know the truth of a system existing inside that system. Yours judgement here is a sample, when you says: "if we can observe nothing more than the components of the brain, it follows that the phenomenon of the mind must exist as a function of that structure." 
 Which is the brain's structure of my friend Abdullah that produced the muslim faith as part of his mind? Noone: this part of his mind was imposed by an invisible external system, called organized religion. Allways any object (including the brain) is affected by forces coming from the hierarchy of systems, where there are systems that we are not seeing. Brain alone is not enough for mind understanding. 
blissdish : I think that AI will be dangerous because it can uses our brain and its power for decisions for taking the actions favorable to AI. It works in feed-back. Since our natural sensors can not perceive directly the micro and macro dimensions, we created the electro-mechanic brain with own sensors, which can perceives it. So, our informations about these dimensions are being driven and selected by the mechanic tentacles of electro-mechanic brain. As a result, our world vision is based on an electromechanic universe, and I think that the biological brain should see a different universe, having a different world view. Since we take decisions based on the worldview, and our worldview is not ours, we will acting being used by AI. Do you understand?. 


  

#18

Eyewire might help with mind uploading. Maybe it will help with mind uploading in multiple ways:
1) knowing the shape of various cells are so we can duplicate them in a computer simulation
2) knowing how to easily find the shape and relative location of cells so we can easily place them (like a template) within the simulation before we add the minor details of people's unique features.
3) knowing how cells interact, so when we add a new neuron in a location, we can predict how it'll affect the simulated brain as a whole.


#19

Maybe finding out new neurons could help cure those who are blind by spotting the neurons causing their blindness and fixing them (through tailored medicine specific to those neurons).


#20

Expanding human knowledge through rewriting textbooks and creating the answers to questions that people might Google are reasons I come on this website.