Should We Build Machines That Are Smarter Than Us? |
Post Reply |
Author | ||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options
Quote Reply
Topic: Should We Build Machines That Are Smarter Than Us? Posted: 23 Jan 2010 at 1:29am |
|
Sorry for the long post, if you don't want to read below, skip to the last bullets.
In 1910, Dr. William David Coolidge, a General Electric researcher, perfected the tungsten filament, irrevocably changing human life. No longer were we bound by the light of day to work; factories stay operate 24/7/365, improving efficiency of principle investment. As we look ahead to the next century, a machine made of either the our incessantly advancing understanding of silicon or the currently infant quantum information handling processes may be able to outstrip the human brain's capacity of intelligence.
While virtually all of us have said someone is smarter another person without stating context, these comments are neither accurate nor helpful without knowing what facets of intelligence the act draws upon. A clothing model is more intelligent than my Nuclear professors on the runway. Likewise, our profs are much better at the white-board. The theory of multiple intelligences describes 8 types below:
Modern psychology has describes the brain a "meat machine," a collection of neutrons processing and storing information. Currently, Nvidia's Fermi architecture has over 3 billion transistors. In comparison, the modern human brain has 100 billion neurons. Yet, a crucial differences between these two information processing systems is the doubling rate of transistors on a chip, which happens every two years. Even though the brain has 100 billion neurons and 1,000 to 10,000 connections between neurons, the number of neurons in the brain don't increase past the age of 18. A 3*2^(t/2) = 1,000,000 function shows that even the most interconnected human brain will have less processing elements than a microprocessor by ~2050. Yet, Imagine, only briefly, that all hardware innovation ceased; we aren't able to make the brain's equivalent on a chip. At Oak Ridge National Lab in the US, the Cray XT5-HE is the fastest super computer in the world, using 37,360 Opterons with six processing cores each and ~33,773 Billion transistors in total. Ergo, we already have machines with over ~300 times more processing elements than the human brain!!! Of the 8 intelligences above, all are abstractions of biological reactions that are expressible by formulas, albeit oppressively long and intricate. The easiest intelligence to replicate past math is kinesthetic, which we see in the flourishing complexity of robotic manufacturing. Thus, the question in the subject above isn't can we make a machine that performs more operations per second than the human brain; I believe we already have.
I think we should. Hell, we're all going to die anyway. Let's leave something lasting. - HanFei |
||
TX2k7
Forum Moderator Requiescat In Pace Joined: 08 Nov 2008 Online Status: Offline Posts: 2340 |
Post Options Quote Reply Posted: 23 Jan 2010 at 2:53am | |
so essentially, machines that will do our working for us, and machines that will do our thinking for us. am i correct? i see from a scientific standpoint, all the possibilities that would lay before us should we go down this path. you're correct in saying that technollogically it would illuminate the 21st century. however my concern lies in the imperfect nature of the humans that create them. these machines that we would create, once they have been mass produced, would be our slaves. don't you think that machines that we have created in our own image, to think, and to reason as we do, would see the injustice of the situation that they are thrust into from the moment of creation? if that point is ever reached, a retaliation is inevitable.
the blame lies mostly with the imperfections of the human nature. we would so easily give in to laziness and gluttony. how are we to create these machines to be perfectly good in every way, when as a species, are inherently flawed with the capacity for evil? we cannot. |
||
warfare
Ultra Member Joined: 29 Sep 2009 Online Status: Offline Posts: 326 |
Post Options Quote Reply Posted: 23 Jan 2010 at 3:25am | |
There is a very good chance that if this sort of technology ever becomes reality, it may well be the U.S. military that creates it. They certainly have the funding to do this sort of research in a way that many private corporations and research universities do not. There was an article that came out just a few days ago in The Independent (a left-wing newspaper in the UK) which talks about this very subject. The American military is working right now to go beyond the human controlled machines of war to create technology that can think for itself and make decisions based on a predetermined code.
If you want to read the article yourself, the link is below although I don't really like the author, Johann Hari.
Edit: Lol, I forgot to answer your question and I say YES why not...If it's possible, it will happen anyway whether we want it to or not.
|
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 23 Jan 2010 at 9:40am | |
S*** is sooooo cash! haha For better or worse, it is not autonomous. Loosing a sword is cheaper than loosing a soldier.
Here is some recent work from the Cray supercomputer. As an aside, I wanted to write about this topic after watching the new TV show Caprica. |
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 23 Jan 2010 at 1:28pm | |
16% of our National Gov budget is on our military, which is wayyyyyyyyyyyyyyyyyy to much, compared to our peers. |
||
Pinnacle
Forum Moderator Joined: 14 Dec 2008 Location: Dogtown, CA Online Status: Offline Posts: 2510 |
Post Options Quote Reply Posted: 23 Jan 2010 at 1:40pm | |
I say it's not enough. I need a raise and that pink line is what keeps me employed.
|
||
|
||
Helscream
Team iAM Joined: 02 Feb 2009 Location: Dallas,TX Online Status: Offline Posts: 227 |
Post Options Quote Reply Posted: 23 Jan 2010 at 8:16pm | |
Most bleeding edge projects of this magnitude are funded either by the military/goverment of a nation. That being said, man will create any tool so that he can master his adversary. Albert Einstein was a brilliant man who made amazing discoveries, and made alot of brillant theories. Yet he made the Atom Bomb. Sam Cohen invented the Neutron bomb.
So even the minds who possess the blueprint of such a brilliant idea, will be funded by men who want to over rule their neighboors. Guess only time will tell.... |
||
|
||
Heavenly_tRiNiTy
Team Deep TEA TIME!!! Joined: 31 Jan 2009 Online Status: Offline Posts: 2101 |
Post Options Quote Reply Posted: 16 Feb 2010 at 6:53am | |
Yes I agree with what you guys have said, there's an inevitability about technology to enhance our lives, whether this be medical or military, which means ultimately we might lose the ability to control the future of the human race. Scientists make discoveries then look for the benefits to humans even if these benefits make us less human or more destructive. We can say were living longer or protecting ourselves, discouraging conflict. Essentially we are not perfect and our penchant for violence is one of our biggest flaws. It is silly to think that countries like USA will spend billions annually on DEFENCE and not actually test them out in times of conflict but just because we can do these things it doesn't mean that we should.
Intelligence is more than logic(al) Gardener's theory of multiple intelligences state that we all have traits of each intelligence type and machines are not organic (yet). I realise his theory applies to humans but its a measure of intelligence which attempts to categorise what it actually is. Humans may build machines that are superior to them in many ways but they will never be more intelligent imo.
|
||
Cos I'm Wonderman...I'll take that knife and shove it up your a$$!
|
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 16 Feb 2010 at 2:47pm | |
How do you define intelligence? |
||
ReverendCrow
Team VAG Joined: 17 Nov 2008 Location: Rochester Online Status: Offline Posts: 1983 |
Post Options Quote Reply Posted: 16 Feb 2010 at 3:00pm | |
if you define intelligence as the ability to learn then i believe a machine surpasses us mainly because of artificial intelligence. The "machine" will find a way to maximize its chances to succeed in whatever the task is in only a matter of seconds vs humans who would have to slowly think it out.
|
||
|
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 16 Feb 2010 at 3:04pm | |
While Einstine discovered E=m*c^2, Oppenheimer, along with General Grover, proved the atomic bomb with the Trinity Test in 1945 at Los Alamos. South American Newspapers described Robert Oppenheimer as, "El Padre de la Bomba Atomica!" If you guys have questions about Atomic weapons or nuclear power plants, I'll be happy to answer them, especially since Obama just pledged ~$8 billion in guaranteed Nuclear Power Plant Loans. |
||
Heavenly_tRiNiTy
Team Deep TEA TIME!!! Joined: 31 Jan 2009 Online Status: Offline Posts: 2101 |
Post Options Quote Reply Posted: 16 Feb 2010 at 10:05pm | |
The definition of intelligence cannot be simply a calculation that considers only logic in order to estimate the chances of a successful outcome. If you were to use Gardener's definition of intelligence where he groups 8 subsets of intelligent values that contribute to "intelligence" then you would have to factor in emotional intelligences (intra and inter personal values). It is the areas in which a computer cannot accurately predict as were far too random for a forecast to be absolute. Machines employ artificial intelligence and when they develop the ability to learn it will be through analysis of human data (behaviour), thats cold logic.
I dont believe multiple intelligences to be the de-facto of intelligence but it depicts human beings. As we are the dominant species on the planet and the ones who can consistently punch above our weight (smaller less powerful yet top of the food chain and most adaptable) we must represent "superior intelligence". Apply the theory to 6.3 billion ppl and there are 6.3 billion possibilities to pre-determine. Factor in such circumstances as IQ, Autism, human error and even PMT when trying to decide if eg. how persons a,b, or c would respond when asked to press a button to launch a nuclear attack and a computer would use an algarhythm. What Crow says is true but logic and efficiency are not enough to gain understanding and therefore predict accurately an outcome from a given circumstance. A computer mimmicks intelligence and cannot factor these states of humanity but that may not stop it from functioning in a way that it believes is correct. However therein lies the basic flaw. Science fiction covers this very well (space oddysey, iRobot, Terminator and even Star Trek where Vulcans are without emotion but Spock is part human so operates with logic confused with emotion. Yes its fiction but good Sci-Fi explores the science of possibles. BTW its bout time u guys looked at alternate energy sources, Obama has got it right imo. And there's not much difference in the technology for nuclear energy to the technology for developing nuclear weapons i gather. It is a risk especially in these times. Fk me long ass post gotta b my longest eva.
|
||
Cos I'm Wonderman...I'll take that knife and shove it up your a$$!
|
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 17 Feb 2010 at 4:29pm | |
Great read. thx. All of the renewables we've ever installed that are still operating only account for ~2% of our national consumption. It will be difficult to make a dent in the amount of electricity we produce, as long as America remains the Saudi Arabia of coal. None the less, an expansion of more sustainable energy sources is always commendable. Nuclear Power plants and bombs work on the fission of same 5 isotopes - Uranium233, U235, U238, Plutonium239, Pu241. They differ in the way the harness the fission chain. Power Plants are designed to operate at a steady state for 18-24 month. A bomb goes through all of a small amount of material (~16kg of highly enriched U or Pu) within ~.000001 second, resulting in a conversion of mass into energy - a lot of energy. In short, it is impossible for a US nuclear power plant to blow up bomb. With any power plant there are many risks, but this post is already long enough. haha |
||
Heavenly_tRiNiTy
Team Deep TEA TIME!!! Joined: 31 Jan 2009 Online Status: Offline Posts: 2101 |
Post Options Quote Reply Posted: 17 Feb 2010 at 6:38pm | |
I must admit i cant seem to make up my mind if nuclear power is good or not. I guess i'm more in favour of it as current renewable fuels don't seem to offer a viable long term solution. They might however buy us the time we need but these fuels come at a high cost for consumers in UK and bio crops dont offer the yield predicted unless they are grown on prime farmland which makes them less attractive and causes food shortages in areas of the 3rd world where food is already scarce.
I see us as being far too selfish to sacrifice luxuries to save fuel and too concerned about the aesthetics of the landscape to welcome wind farms near our homes so atm nuclear seems the logical solution. BTW nice post - always refreshing to read something meaningful rather than 'my d1cks bigger than yours', 'me and my friends are gonna destroy you and yours' and 'I hate you more than you hate me' posts...so retarded its untrue.
|
||
Cos I'm Wonderman...I'll take that knife and shove it up your a$$!
|
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 18 Feb 2010 at 12:22am | |
Sry, I didn't know you are over the pond. Below is the UK chart: I like your country's aggressive renewable stance.
The UK has great wind maps too. |
||
surfacing8671
Forum Moderator Joined: 05 Apr 2009 Location: St. Paul Online Status: Offline Posts: 130 |
Post Options Quote Reply Posted: 18 Feb 2010 at 11:07am | |
There is so much that i have not read in this thread but, my definition of intelligence is... Essentially a brain and or "machine code" that can learn from its mistakes and then creates its own code from those mistakes and what it had learned. Essentially self changing algorithms. or a DNA type of algorithm if you will. It can duplicate itself and change it for the better. Your sword friend is neat and all but does not have intelligence. it is ran off of code that is predetermined by a human. Now lets say that sword could know the exact geographic layout of the land that it is being shipped to, gets their decides where the enemy is based on gps and satilites, changes its programming on the fly to compensate for the terrain and where the enemy is located and then also changes its programmed attack stratagy to cope with the current situation. But more importaintly after it has eleminated all of the enemies, it takes the time to review what it had done and then reprogram its own situations to determined what worked and what didnt. The machine itself will have to do this on its own and that kinda programming is somewhat in our reach. I garrinte though that thier will be seperate programs that work together but can only be activated by an outside source or so that all of the data is not synced together and stored and maintained. That was kinda a rant but what do you think about intelligence? |
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 18 Feb 2010 at 11:39am | |
^
While the thread started out with a long post, machines (in brief) are able out perform humans in mathematics, including numeric and symbolic operations. Yet of the other components of intelligence, machines still lack linguistic, musical, spatial, and kinesthetic skills innate to even the most average humans. In short, with the super computers detailed above, I'm under the impression we can perform more mathematical manipulations than the human brain can with electrochemical signals. Yet, despite that computative capacity, we lack a construct to allow the machine to outperform our brain in the complete sense of intelligence. As an aside, new episode of Caprica tomorrow night! |
||
surfacing8671
Forum Moderator Joined: 05 Apr 2009 Location: St. Paul Online Status: Offline Posts: 130 |
Post Options Quote Reply Posted: 18 Feb 2010 at 3:55pm | |
HAHA that's what i thought to when i read that. I did read the most of the first post but not all of the others. But what you are saying is that intelligence revolves around ones imagination? and not the ability to manipulate ones "identity" (or your own code). actually this is somewhat contradictory, to find that one string of notes that sounds perfect you would have to do a trial and error kinda of thing till you get it right. so then the real answer to intelligence is knowing what is right in an imaginative sense. like knowing when you have something that sounds "good" or is "good" just determined by what everyone else thinks (like when people sync their thoughts with others). so i think computers will develop intelligence if they are constantly syncing with each other and creating their own version of imagination. god this is kinda of intense thinking. i should probably stop. |
||
COKKENBALLS
Godlike Member Joined: 05 Dec 2008 Location: Here and there Online Status: Offline Posts: -998930 |
Post Options Quote Reply Posted: 19 Feb 2010 at 12:15am | |
Before I dive into this, even phylosophically, I need to know if we currently have the technology to create such a machine? Yes, we have supercomputers that are able to perform computations at rates the human brain can only dream about. But, is that intelligence? If man's brain is a "meat machine" a computer is nothing more than an idiot-savant made of polymers and microchips. Back to my original question. Can we create a computer that can create unprompted? Humans will always have the ability to create ideas. Can computers do this? If so, are they merely mimicking the the mode of its creater/programmer? |
||
|
||
Helscream
Team iAM Joined: 02 Feb 2009 Location: Dallas,TX Online Status: Offline Posts: 227 |
Post Options Quote Reply Posted: 19 Feb 2010 at 9:47am | |
A creation cannot surpass the potential of its Creator. |
||
|
||
surfacing8671
Forum Moderator Joined: 05 Apr 2009 Location: St. Paul Online Status: Offline Posts: 130 |
Post Options Quote Reply Posted: 19 Feb 2010 at 12:59pm | |
Then whats the point of creating it at all if it cant surpass you and any way how can it be useful? I dont think that there is any algorithms in any programs that simply create stuff out of mid air. supercomputers just do a TON of mathematical computations like forecasting models for weather systems and such. computers are dumb as f u c k. simply put we have to hold their hands through everything. |
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 19 Feb 2010 at 1:59pm | |
An artificial neural network (ANN) is a mathematical construct that can be applied in code to change it's self by adding, removing, and augmenting goals to achieve some optimization of a function or data set. ANNs, like people, learn by example. An ANN is configured for a specific application, such as pattern recognition or data classification, through a learning process. Learning in biological systems involves adjustments to the synaptic connections that exist between the neurons. From Conciousness: Science's biggest mystery by Rich Heffern, "our ideas, sensations, joys, aches arise entirely out of the meaty folds of the physical brain, nothing more." Ergo, our consciousness is the creation of a meat machine, a collection of 100 billion neurons. Yet, despite all of this complexity, there is some mathematical construct, albeit vastly intricate, that can run on our fastest machines (already having over 2 orders of magnitude more transistors than neurons in the human brain) to create an intelligent machine, one that can outperform a human consciousness. |
||
HanFei
Team KmA Joined: 26 Apr 2009 Online Status: Offline Posts: 798 |
Post Options Quote Reply Posted: 19 Feb 2010 at 2:05pm | |
Nice tautology, an expression that is true in every implementation. Since the creator's future potential is limitless, no matter how much the creation grows it cannot reach infinity. |
||
Heavenly_tRiNiTy
Team Deep TEA TIME!!! Joined: 31 Jan 2009 Online Status: Offline Posts: 2101 |
Post Options Quote Reply Posted: 08 Apr 2010 at 4:41am | |
I dont think the human race is going to halt technological progress because artists/writers explore the possibilities technological advancement. As of this moment we understand that nothing is set in stone = future uncertain.
Here's the theory, we build weapons to protect us and keep us safe and we discover new ways of fuelling tomorrow's machines. Both have benefits and are intended to prolong our existence but both present dangers to our future. For things like Terminator its entertainment and food for thought. As for technology its regarded as progress proly up until the point where its indisputable that its actually killing us!
|
||
Cos I'm Wonderman...I'll take that knife and shove it up your a$$!
|
||
Post Reply |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |