Jump to content

  • Log In with Google      Sign In   
  • Create Account

FREE SOFTWARE GIVEAWAY

We have 4 x Pro Licences (valued at $59 each) for 2d modular animation software Spriter to give away in this Thursday's GDNet Direct email newsletter.


Read more in this forum topic or make sure you're signed up (from the right-hand sidebar on the homepage) and read Thursday's newsletter to get in the running!


human intelligence


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
97 replies to this topic

#61 LennyLen   Crossbones+   -  Reputation: 4026

Like
0Likes
Like

Posted 03 February 2014 - 07:37 AM


Perhaps we should have sorted out the definition of "intelligence" we all can agree on first. 

 

This is a very good idea.  I would say the definition of intelligence is the ability to apply existing knowledge to new situations and to be able to acquire new knowledge.

 

Is just giving a definition enough though?  We can obviously use that definition to show that computers as we know them are not intelligent.  While they can be made to look like they are learning, this is purely as a result of clever programming, using tools such as conditionals and lookup tables.  But what about us?  While we may feel that we are not programmed and that we really are learning from our experiences, we don't know enough about how the brain works to definitively say that any decisions we make are not just a result of predetermined arrangements of neurons.  



Sponsor:

#62 Felix Ungman   Members   -  Reputation: 1066

Like
0Likes
Like

Posted 03 February 2014 - 08:26 AM

 
I think Alex is unto something by defining it as maximizing future freedom of action. A good definition should be based on action and behavior, and should make it possible to measure and compare intelligence, at least in theory.

openwar  - the real-time tactical war-game platform


#63 Tutorial Doctor   Members   -  Reputation: 1692

Like
0Likes
Like

Posted 03 February 2014 - 04:25 PM


I would say the definition of intelligence is the ability to apply existing knowledge to new situations and to be able to acquire new knowledge.

 

I really like that definition, but it seems too short to be accurate. What I mean is that with a little thought, perhaps someone could find a case in which such a definition would not demonstrate intelligence but some other term. 

 

I also like the phrase "maximizing future freedom of action." But again, I think there is also a scenario where such a definition would demonstrate some other term other than intelligence. 

 

When I thought "intelligence" I thought "the ability to gather information." 

 

Now, what term might define the ability to apply information gathered? Or what is knowledge? Or what is wisdom?

 

I always considered wisdom to be the tactical use of knowledge to make the best decision in a given situation. I considered knowledge to be stored information. Information that is not stored would not be a part of your "knowledge database"

 

Yet, one can "forget," and then later "recall" and then that information is retrieved from some dormant part of the brain. So was that knowledge lost or forgotten?

 

Now, I am not being philosophical, but these things seem like they would be important in this conversation. Do computers forget? How much info can they retain? 


They call me the Tutorial Doctor.


#64 LennyLen   Crossbones+   -  Reputation: 4026

Like
0Likes
Like

Posted 03 February 2014 - 06:46 PM


Or what is knowledge?

 

I would describe knowledge as a set of data that describes a specific scenario.

 


I always considered wisdom to be the tactical use of knowledge to make the best decision in a given situation. I considered knowledge to be stored information. Information that is not stored would not be a part of your "knowledge database"

 

I've always thought of wisdom as being an accumulation of knowledge.  If you consider someone like Albert Einstein, it's very evident that he was an extremely intelligent individual.  Now consider him as an infant - while he would have been just as intelligent, he would not have been wise as he did not have enough knowledge to use his intelligence.  Likewise you can have people who know a lot of things but who lack the intelligence to take what they know and apply it to new situations.

 

 



Now, I am not being philosophical, but these things seem like they would be important in this conversation. Do computers forget? How much info can they retain? 

 

I don't think the ability to recall information should be considered intelligence, though it does obviously affect the decisions an intelligence will make.



#65 Felix Ungman   Members   -  Reputation: 1066

Like
0Likes
Like

Posted 04 February 2014 - 01:40 AM

Finding a precise definition that completely captures the concept of general intelligence will be as hard as finding a precise definition of beauty. It might be interesting from a philosophical standpoint, but from an engineering standpoint, narrow definitions are usually easier to work with, even if the only capture limited aspects of the underlying concept. It's perfectly fine to have several definitions or tests of intelligence and realize they capture some but not all aspects.

 

The problem with defining intelligence in terms of knowledge or wisdom is that you should be able to determine intelligence with black-box testing. Internal structure shouldn't matter.


openwar  - the real-time tactical war-game platform


#66 Tournicoti   Prime Members   -  Reputation: 684

Like
0Likes
Like

Posted 04 February 2014 - 08:45 AM

I think this topic is flawed from the very beginning, because each one of us has his own definition about what human intelligence is.

For some, it's the ability (the state), of being able to accomplish a given task (playing chess, whatever). For some others, it is the ability to learn something ... (being able given DOF, to learn to walk for instance) . A function.

 

Before answearing the OP question, we have to be on the same wave about what intelligence is. And this is a philosophical and still open question ...


Edited by Tournicoti, 04 February 2014 - 09:07 AM.


#67 Tutorial Doctor   Members   -  Reputation: 1692

Like
1Likes
Like

Posted 04 February 2014 - 09:56 AM

Right tournicoti. Even looking at the entry on Wikipedia, leaves it rather open ended.

I had just considered "overall" at first, and I do think that "overall" human intelligence is far more advanced, mainly because of the many diverse situations it can handle, even from birth, autonomously.

The reason I said a computer would have to be able to debug itself in order to be considered close to our intelligence is that we can solve new and foreign problems a lot easier using non-standard ideas. We can create new plans for unfamiliar situations etc...

It is sorta strange though that so many things have open ended definitions, perhaps because definitions change over time, but that is contrary to the definition of "define" which is for the purpose of defining something as being different from something else, so then we have to ask the questions,

What is intelligence?
What is knowledge?
What is Wisdom?
What is forgetting?
What is recalling?
What is remembering?
What is ignorance?
What is uneducated?
What is educated?
What is assumption?

Etc.
Are they all the same? No. What defines each term as different from the other. Even in this simple scenario, computers just aren't there.

They call me the Tutorial Doctor.


#68 samoth   Crossbones+   -  Reputation: 5037

Like
0Likes
Like

Posted 04 February 2014 - 10:36 AM

I would summarize it roughly like this:
  • Understand causalities.
    • When I send out this nerve impulse, my head turns.
    • When my head moves, funny signals come from my inner ear. These are always the same for identical moves. Also, the world always moves in the opposite direction.
    • When I drop my teacup, it falls to the ground. When I drop my cell phone, it falls to the ground.
  • Combine related and unrelated causalities, and make decisions based on that combined knowledge.
    • Most things I tried, including teacups and cellphones, fell to the ground when I dropped them. I expect a sausage to fall to the ground as well, without having to try.
    • Often, things break or become dirty or otherwise unusable when they fall to the ground.
    • I don't like that happening (also my back hurts from picking them up) and therefore will not voluntarily drop things.
  • Perform sanity checks on decisions before and while taking action.
    • The GPS tells me to turn right. Into the river. Screw that cheap piece of junk.
    • The traffic light just switched to "Don't Walk". I don't think that I'm supposed to stop right here in the middle of the street.
  • Validate decisions, as their outcome becomes apparent.
    • I really should have dropped that darn cellphone and reached for the escape ladder instead.
    • Giving that tramp a ride was a stupid idea.
  • Improve future decisions based on the outcome validation.
    • Last time I got beaten up bad after pushing that guy. This time I'll pretend to give in and kick him in the balls from behind.
    • Next time someone asks you if you're a god, you say "Yes!".
  • Do all of the above without someone having to tell you, and without someone telling you how.
  • Adjust decision parameters by observing the outcomes of other people's decisions.


#69 Álvaro   Crossbones+   -  Reputation: 13933

Like
0Likes
Like

Posted 04 February 2014 - 01:04 PM

It is also possible to skip most of the philosophical aspects of the matter by realizing that compression is very closely related to intelligence. I know it sounds odd, but it is a deep and interesting idea.



#70 samoth   Crossbones+   -  Reputation: 5037

Like
0Likes
Like

Posted 05 February 2014 - 03:41 AM

It is also possible to skip most of the philosophical aspects of the matter by realizing that compression is very closely related to intelligence. I know it sounds odd, but it is a deep and interesting idea.

Although it may seem that way, I think the idea is somewhat flawed. What compressors do is reduce entropy. If they work well, that is.

 

Dictionary-based compressors replace parts of the input with things they have stored in their dictionary, if that takes fewer bits, according to a scheme that the programmer has decided. Maybe the programmer is intelligent, but the program doesn't do anything special.

 

Statistical compressors, on the other hand, determine the likelihood of the next symbol that comes in by what they have seen before. You could much more easily associate this with "intelligence" even if they still only follow a very rigid pattern that the human programmer has devised.

 

But if you think about it, this is a behavior which is very much akin to a gambler playing roulette, who, upon having seen 17 three times, puts his money on 17. After all, 17 seems to be the lucky number. Would you deem this intelligent? Secret tip: A horse with a name like "Lightning" cannot lose.

Maybe PPMD uses a somewhat more sophisticated algorithm, but in the end it is the exact same thing. Looking into the crystal ball.

 

The only difference (and the difference that decides on the outcome!) to compressing enwik8 is that the input is different. Characters in enwik8 are not random, but are highly correlated, and there is a huge amount of entropy in that text. This is why compressors are more successful than our roulette gambler. Still, they do more or less the same thing. They use some statistical model and if they are lucky, this allows them to correctly predict the coming symbols.

 

What would be more impressive would be compression in a sense such as you show an image to a computer and it outputs  "fat guy in funny clothes making a sad face, bent over a dead girl, that's Pavarotti as Rigoletto". That is much more like the way a human would "compress" that photo.



#71 Paragon123   Members   -  Reputation: 473

Like
0Likes
Like

Posted 05 February 2014 - 04:36 PM

How is it possible that a human brain is not a finite-state machine? How can the human brain possibly work any differently at an atomic/cellular/electronic level then a computer does? 

 

Genetic Algorithms prove that a computer can recalculate possible solutions in a manner that helps arrive at a correct conclusion. Further, the fact that humans can sometimes be stumped is very analogous to a genetic algorithm that was unable to successfully arrive at a correct solution.

 

A couple years may be too soon... but saying that we won't have the power to do it in 100 years is like when  it was said that no one will need more that 637kB for a personal computer. 

 

A computer can perform calculations faster than MOST human brains... however there are people who have beaten computers at calculating ridiculous numeric calculations (http://en.wikipedia.org/wiki/Shakuntala_Devi). In fact, I believe that if the human brain where "programmed" to do calculations it would out perform current computers easily... but the difference being that the "thinking" algorithm we have is not optimized for numeric calculations, but instead it is optimized for learning, recalling, idea association, and short cut finding/adjusting. Further, I believe that the first true A.I will have intelligence on par to that of the original single celled organism... barely anything that would actually be considered A.I by "SciFi" Standards and it will eventually evolve it's self to higher and higher levels of intelligence until it eventually surpasses that of humans.. Although, I suppose, it may also be possible that humans are already "learning" and gaining intelligence at the "most optimized" speed so perhaps A.I will always be X steps behind. 

I think that thinking some one will eventually write some code, hit F5, and see... "Hello Dave, I think therefore I Am." is extremely far fetched... Infants take a month to learn how to roll over... and a year+ to even say a single word that has any relevance to exterior factors...  It seems pretty likely that A.I will model True intelligence and I think initial the initial learning phase (An infant is not at the initial learning phase, evolution has been working on it for millenia) is one that won't be able to be skipped. 



#72 Álvaro   Crossbones+   -  Reputation: 13933

Like
1Likes
Like

Posted 05 February 2014 - 04:53 PM

A computer can perform calculations faster than MOST human brains... however there are people who have beaten computers at calculating ridiculous numeric calculations


That statement is absurd. Shakuntala Devi took 28 seconds to compute 7686369774870 * 2465099745779. That's quite a feat, no doubt. It would be hard for me to measure how long it takes my laptop to make that computation, but the order of magnitude is 0.00000001 seconds.

#73 Paragon123   Members   -  Reputation: 473

Like
0Likes
Like

Posted 05 February 2014 - 06:37 PM

 

 

Examples of the problems presented to Devi included calculating the cube root of 61,629,875, and the seventh root of 170,859,375.[3][4] Jensen reported that Devi provided the solution to the aforementioned problems (395 and 15, respectively) before Jensen could copy them down in his notebook

 

Right, I suppose it was an exaggeration to claim she beat a computer, but the point was that if a human brain were to be wired to be used exclusively for calculations, I'm sure it could beat a computer.



#74 Álvaro   Crossbones+   -  Reputation: 13933

Like
2Likes
Like

Posted 05 February 2014 - 06:45 PM

Right, I suppose it was an exaggeration to claim she beat a computer, but the point was that if a human brain were to be wired to be used exclusively for calculations, I'm sure it could beat a computer.


Well, if a computer processor were to be wired to think, I'm sure it could beat a human. That makes about as much sense.

#75 Paragon123   Members   -  Reputation: 473

Like
0Likes
Like

Posted 05 February 2014 - 09:15 PM

Which was exactly my argument.... I was saying precisely that a human brain is no more than an extremely powerful computer. Our brains are made of the same matter and use the same electricity, it stands to reason that either, 1) A synthetic brain could be designed to do the same thing more efficiently or 2) Our brains are completely optimized such that a synthetic brain could at best match our thinking ability. 



#76 Tournicoti   Prime Members   -  Reputation: 684

Like
0Likes
Like

Posted 06 February 2014 - 01:43 PM

Comparing the computational power of a "laptop" to a human brain makes no sense. This is irrelevant.This is even absurd. Train your laptop to speak... you''ll get your answer

Computers are good to execute fastly algorithms humans order them to do, that's it. Talk to your laptop ... and ask it what intelligence is, you'll be very disappointed about its response ... Anyway, the problem is still the same : WHAT IS INTELLIGENCE ? This is not a stupid phylosophical question (like if it was out of the field of this 'discussion" ...., and an easy and unfair  way to skip the question ). This is the CORE of this (stupid) discussion


Edited by Tournicoti, 06 February 2014 - 02:50 PM.


#77 Tournicoti   Prime Members   -  Reputation: 684

Like
0Likes
Like

Posted 06 February 2014 - 02:23 PM

I do think we will be able to reproduce artificially intelligence, even better intelligence, but this has nothing to do with decision tree or negamax or whatever, ... it would anyway be the product of a reverse-engineering that we will not completely understand ...


Edited by Tournicoti, 06 February 2014 - 02:38 PM.


#78 Grimshaw   Members   -  Reputation: 657

Like
0Likes
Like

Posted 06 February 2014 - 03:07 PM

From every source I every read, watched or been told about, it seems the humanity didn't figure out how the brain works entirely. 

 

Apparently, the mankind, over many years and experiments as discovered a few things about the human brain. Enough to help with psychology, neurology and even create cool gadgets, but we don't seem to be even close to fully understand our conscience. There are theories claiming it is a quantum computer, in which case we probably can't replicate it with any technology that resembles what we have currently.

 

To sum it up, its not a matter of processing power. Even if we had infinite computational power, we still don't know how to replicate human brain as a whole. Any A.I that seems human is merely a incomplete emulation. An approximation. We don't have or will have anytime soon enough power to simulate all those neurons in a computer program, and every approximation we can make is amazingly good at computing numbers but amazingly bad at having anything that feels like conscience.

 

If there is something wrong with my post let me know :)


Indie Game Developer - Daeva Theory Studios

 

Twitter - Facebook

 

Next

Terra - rise of mankind

 

Previous

SFML Game Development (Book)


#79 Álvaro   Crossbones+   -  Reputation: 13933

Like
1Likes
Like

Posted 06 February 2014 - 03:29 PM

Grimshaw, my main objection to what you wrote is that we are talking about surpassing human intelligence, and replicating how a human brain works is only one possible way to do it. But a more promising way to do it is to concentrate in tasks where currently humans are better than machines (e.g., cooking) and try to make machines that do better at those tasks, possibly using completely different mechanisms than humans.

 

We have machines that can go very fast, but they don't move on legs like animals do; we have machines that fly, but they don't flap their wings the way birds do. I think we can build machines that are good at many tasks that one would consider part of intelligence, and we don't need to worry about what consciousness is or any other nonsense.

 

In related news, I am only interested in solving problems where progress can be measured. Consciousness is something that doesn't interest me much.



#80 Paragon123   Members   -  Reputation: 473

Like
1Likes
Like

Posted 06 February 2014 - 04:39 PM

I'm trying to think of a good analogy because I think I am being misconstrued. 

The first point I am trying to make is that Both computers and Brains are physical objects, each of them are powered by electricity and make calculations based on the flow of that electricity through their components. At a fundamental level it is impossible that a synthetic brain can not be created as their is no magical, or impossible element a brain possesses that can not be simulated.

 

The next point Is that pure computational power does not "cause" intelligence, it is the application of that computational power from which intelligence emerges. A Lamborghini has more Horse power than a tractor trailer, but the tractor trailer applies it's horse power to be able to pull much heavier loads (through torque) than the lamborghini does. While this statement can indicate that a brain actually processes data slower, it is my opinion that a human brain has more raw processing power than current computers.

 

Comparing the speed at which an action is done between two machines is not indicative of which machine is working faster. For example, if you wanted to retrieve data from a database you have a multitude of options. Suppose one machine does a simple "Select X,Y,Z from TBL" while another uses a complex and well designed ORM. The first, naive implementation will execute more quickly simply due to the reduced number of instructions that need to be processed, but the ORM version will be much more flexible and complete.

 

The "intelligence" algorithm, will be/is orders of magnitude more complex than any algorithm a computer has ever been given, which is why a computer can complete particular problems more quickly than a human... but the "intelligence" algorithm is still nothing more than an algorithm, which when properly understood will be just as easily followed by a synthetic brain (computer) as a human brain. 

 

 

The comparison of computational power between the human brain and a computer was meant to illustrate the following:

 If the computer is running the simple select, and the brain is running the ORM and the brain is outputting the solutions more slowly then the computer... then I believe that the human brain running the simple select would easily outpace the computer...

 

This doesn't rely on a definition of "what is intelligence"... no matter how it is defined, it will ultimately be an algorithm, which a computer will be able to run.  

 

You say that a computer can only do what it is told to do, and while I completely agree, I also believe that there exists a method in which we could tell a computer to learn, comprehend, imagine, innovate and create. And I believe this because i think that at the most fundamental way our brains must work as some form of computational device that is running a series of definable algorithms.






Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS