Is there a limit on the intelligence of created beings

Question

I'm looking for arguments/theories regarding this question:

How far can a created being advance intellectually relative to the creator/environment.

Let me elaborate:

suppose there are some very intelligent programmers dedicated to engineer the most advanced piece of program, how advanced can the program be? Can it understand the hardware which it runs on? Is it possible for the program to comprehend the motives of its creator? Also, will it able to understand the physical world that the computer exists within?

There are some physics theories that treats the universe as a finite state machine. If that's the case, one can argue that intelligence in the state machine would have to be a subset of all possible states, and no intelligence would be able to comprehend supersets of all possible states, namely the container universe of this universe.

Perhaps humans are in a similar situation, since we are part of this universe, maybe it is impossible for us to understand the basis of the existence of this world, similar to how a program would know nothing about the world outside of the computer hardware unless the programmer supplies some information.

Is it then pointless to ask the question "where do we come from"?

Answer 1

I don't know of any specific philosophers that have written about this, but a few thoughts:

1) Are there limits on intelligence? This seems to be the case. We live in a universe that had a distinct creation time, and which will only last a finite amount of time. Further, there are a finite number of atoms, which prevents the possibility of a created intelligent being of finite life but infinite size. So, any created being has the obvious upper bound constraint of all of time and space.

For any single being, the constraint is significantly shorter - a single person cannot use all of time and space, he or she can use at most his or her lifetime. One interesting aspect of your question is that "a created being" cannot get very far in understanding its environment - no one person has gotten very far at understanding even how to make a pencil. It is easy to forget, but much of our understanding and ability to manipulate things is due not to individual intelligence but network effects of many people.

2) Can it understand the hardware which it runs on? Yes, at least in some sense. Software can emulate/virtualize hardware. However, trying to extrapolate this to the point of your analogy, this doesn't seem knowable with respect to if we can understand the underlying "hardware" of the universe if it turns out there is something beyond the material universe.

3) Is it possible for the program to comprehend the motives of its creator? The point of this analogy is whether humans could possibly understand the motives of some Infinite Creator or something along those lines. The analogy strikes me as interesting but of limited use: if we created a robot or computer program with some self-awareness, it would be finite just as we are. However, an Infinite Creator is infinite where we are finite. So it seems possible to build a program that understands our motives and still not be able to understand an Infinite Creator's.

4) Will it able to understand the physical world that the computer exists within? This seems very possible. It would only require sensors for detecting things in the physical world (which is easy), a method for sampling the physical world (harder, but still doable), and a way of taking observations and constructing a model of the physical world (harder still, but still possible).

By analogy, is it possible for us to understand something of the "container of this universe?" I think your analogy here is good - a computer program can only begin to understand things in the physical world for which it has sensors. Likewise, we could theoretically only understand a world beyond us to the extent we have been given the ability to sense anything about it.

5) Is it then pointless to ask the question "where do we come from"? It seems that approaches to this question can be put into two groups:

1. We can use reason to understand something of it. (e.g. Descartes)
2. Reason can help, but we first must be given some help from whoever-it-is-that created us.

Approach #2 seems more common today - most people try to take their lead from data and extrapolate from that - atheists say that we have been given no evidence for anything beyond this, so that's the end of the story. Most religious groups believe that God, through revelation, has given us the needed data that supports belief in something beyond. In any case, the question itself seems useful, but depending on your perspective it might be less (for atheists) or more (for religious people) interesting.

Answer 2

Your question first brings up the eternal question of what is intelligence?

Is intelligence simply knowing something, in the case of a program, storing the information in memory? Or does the program have to comprehend and understand the data it has in memory. To what degree does the program have to understand the information to be considered intelligent?

I think it would certainly be possibly to create a program that can perceive and store information beyond what a human can. We do this now with systems like the LHC. Humans have no way of even perceiving most of the information the LHC generates, but we can obviously understand and interpret the information better, and with modern databases, programs can obviously store more information. I would say neither perceiving nor storing information defines intelligence though.

At this point there is not a program created that "understands" anything. All current computer programs simply take a path through their code resulting from input received.

This brings up the issue of free will and how it relates to intelligence. Is free will required for intelligence? Is free will synonymous with intelligence? Do humans even have free will, or are we just programs ourselves responding to inputs?

My personal opinion is that programs will one day surpass humans in intelligence capability. My reasoning behind it is thus:

1. Computers are already much faster than humans in term of data processing ability and the distance is only growing. E.g. Computers can "read" and interpret a large database much quicker than a human could by simply scanning the entries with their eyes.

2. We have such things as computers that can "learn" and adapt their own structure. We also have programmed many evolutionary algorithms that rapidly evolve code to optimize certain parameters. (see http://vimeo.com/79098420)

3. The main thing missing in my opinion is time. I can't imagine a world 100 years from now, providing technology advances at the same rate, where we don't have machines that are as "intelligent" as humans. From there it is another matter of time. Robots and machines can travel long distances in the universe without worrying about time being a factor. Making robots that can create newer and smarter robots from their experiences exploring the universe, (they could find raw materials throughout the universe) would allow humans to explore the entire universe by proxy.

Physical systems that treat the universe as a finite state machine may be correct, and in that case it may not be possible to "understand" the universe. That doesn't preclude the possibility of humans creating something more intelligent than themselves.

I don't believe any question is ever pointless, "where do we come from?" least of all. Since the beginning of humanity this questions has driven people to research and expand their knowledge in pursuit of the answer, and through that we have discovered much about our universe. Instead of being a pointless question I think it is probably the most important question even if it is unanswerable.

Answer 3

not sure what you're getting at.

Can it understand the hardware which it runs on?

are you asking about self awareness? if so i think: yes.

Is it possible for the program to comprehend the motives of its creator?

why wouldn't it? if motives are out there in the world the computer inhabits as much as anything. as long as motivations can be comprehended, i don't see why a computer programmed to do so couldn't process human motivation. when a game engine correctly guesses that you are going to go pick up the ammo, isn't that comprehending a motivation. unless, again, you are asking if they can do so with self awareness.

Also, will it able to understand the physical world that the computer exists within?

again - do you mean sentience?

Perhaps humans are in a similar situation, since we are part of this universe, maybe it is impossible for us to understand the basis of the existence of this world, similar to how a program would know nothing about the world outside of the computer hardware unless the programmer supplies some information.

but then, humans are self aware. we may not ever be able to know some things, ether due to limited time, intelligence*, or just because somethings have no answer to "why" - despite being reasonable questions. there are different kinds of why questions [see salmon on the philosophy of science] and some questions ask "why" in a way that asks for consolation or justification; you may agree that some people cannot be consoled.

Is it then pointless to ask the question "where do we come from"?

it depends on what you are asking! it's useful to know who your parents are, e.g..

• i think intelligence and not just time is a limit to what we can know. a calculator is not ever going to be self aware [without some serious reprogramming!], just like i am not ever going to be able to tell you pi to 1,000,000 decimal places.