Discussion: Douglas Hofstadter

Pages: 1 2
Read "D" as delta... this is an equation I was tought in Biology at university:

DG = DH - DS/ i

were DG is Energy free from Gibbs, DH is Energy available, DS is Entropy an i is information. Why did we see this in Biology? Complex organisms are built with more information and so have more capacity to use energy in a more efficient way. More information reduces the energy lost as Entropy (DS/ i). Imagine an organism that manages infinite information. The term (DS/ i) would become 0 and the Energy it uses would equal the Energy available, although I don`t beleive on it, GOD.

>>By rafo


To mg. Telling oneself you would like to eat ice while nobody is around to hear it. I would like to say we are speaking a language to ourselves, or that language builds itself on thought, since we are taught language as babies. I have to mull this one over.
To rafo: I have always questioned the concept of Entropy. I need to think this one over. But please give me more of your thoughts on this one. I am interested in what you have to say.

>>By redcatdave


Entropy is a measurement of inefficiency. When energy is transfered, some of it is lost (as heat, sound, etc) in the process. Sometimes entropy is referred to as a measurement of disorder (for example, how disordered are the molecules of a substance). I have just found this page in the web, we should try it:

http://www.2ndlaw.com/

>>By rafo


I'll stick to 'the definition of language'...

Rafo,

> The algorithm maybe controls itself, but within the parameters programmed.

So you want to say, that the definition of 'thinking' includes 'not being restricted to programmed parameters' or 'not being restricted at all' ?

> I think the advantage that we humans have is we dispose of 5 perception channels, and we use them all at
> all times

But a standard computer these days has even more perception channels. The network-card, an usb-connector, the keyboard, the mouse, the line-in of the soundcard, an infrared sensor... It receives input on all of these all the time and acts accordingly.

> (watch the computer monitor and your brain will "know" how it would taste, smell, feel and sound -or not sound).

Hmm.. It looks like my computer senses the current situation in a similar way. I have 'wake on lan' turned of. That means when it goes into suspension mode (after not receiving keyboard input for a while) it will turn of the harddrive and other components. It also stops listening on the lan (network). It thinks 'no typing since 10 minutes ? Probably no important data will arrive over the net as well'.
Is there a qualitative difference between 'no speakers -> no sound' and 'no keyboard -> no lan' ?

By the way, welcome to page two of this thread. 3 people knotting an eternal golden braid...

>>By mg


To Mg:
Yes, I know what you mean and it has sense. We could say that the algorithm thinks, but it only thinks within that level (the level for which it is programmed). We could say that it META[processes data]... but that's it. It won't learn any further. It would be "one loop" learning. I think we humans constantly learn from experiencing new things every time (and that changes every time our validation of truth and so we have to adjust our thoughts). So once we META[processed data], we jump to META[META[processing data]] which is something I doubt computers do. Fortunately for computers, they manage a pretty absolute truth. They also don't have to live up and interact with other members in the computer society.
About the perception channels, I don't know to what extent computers can overlap the data they pereceive into one same data base which we could call "experience" and therefore associate for example, the fact that it senses (or not senses) you typing in the keyboard with not seeing you around.
I have just realized: you might be thinking that computers DO interact with other computers, but they all manage the same truth because "god" (read it bill gates) programmed them all. What would have happened if 10 years ago we left a PC and a Mac to interact with no compatible programme? Would they by their own could have learned to live toghether?

>>By rafo


To rafo:
Before I go back to the subject of language, I have to confess I have never believed in entropy. We know what the equation is, but that is a human equation and perception.

To both mg and rafo:
You both have some very interesting thoughts on language and computers. What would happen if Bill Gate the "god" had remained out of the picture? And where would the idea of "thinking" come in? mg, I like your definition of thinking, "not being restricted to programmed perameters." Could computers honestly reach the point in which they "think" on their own accord?

>>By redcatdave


> I like your definition of thinking, "not being restricted to programmed perameters."

Oh no! That's not my definition of thinking. I just asked rafo if this is an aspect of his definition of thinking.

So far, my definition of thinking is still 'processing data'. Processing of data probably already kicked in with the big bang. I don't think Bill Gates had a big impact on this.

>>By mg


The comment about Bill Gates being "god" was a simil... I want to put that clear because if I was to believe in god, it certainly won't be even close to Bill Gates.
Mg: I know it is not your definition, but I think it's good ("not being restricted to programmed parameters"). Because we humans did not create the world, we constantly have to deal with new information, new data, new truths. So thinking would be a process of adaptation: finding a way of matching new truths with the ones we already manage. So I think it can't just be processing data because what happens with new data? How do we know how to process it? For example, what happens if in the Calculator application we write the number in letters instead of typing the digits?
Redcatdave: Mmm... I do think that entropy is useful as an abstract concept, more than being a formula. I think everything has a degree of inefficiency: happiness, our organism, chemical processes, sports, art... and learning, gaining more information is a way to reduce those inefficiencys.

We are starting to mention the creation of the universe...there is something that hangs me up, what would be a good reason for creating the universe? What is the function of it?

>>By rafo


"not being restricted to programmed parameters"

Both of you seem to be attracted to this being part of the definition of thinking...

I agree, that the funtion "ProduceOneMillionPrimeNumbers" that I wrote above is restricted and can only choose between 2 approaches to solve the given problem. That is because its only purpose was to show that even a computer can monitor and optimize itself. At that point, you said computers don't 'think' because they don't monitor and optimize themselves. Now you say, computers don't think because they are restricted to given parameters.

Let's look at another function that has more degrees of freedom:

function SolveTheProblemBruteForce()
{
.while(ProblemSolved==false)
.{
..TestProg=CreateARandomProgramm();
..Execute(TestProg);
..ProblemSolved=TestIfTheProblemIsSolved();
.}
.print ("problem solved!");
}

In natural language: This function will try one random approach after another until a given problem is solved.

This algorithm can produce any imaginable aproach to solve the given problem and can handle any kind of problem. Looks like it's not restricted at all.

So, does this algorithm meet your definition of 'thinking' ?

>>By mg


Hahaha! Rafo, I like your concept of Bill Gates. I tend to regard him as more of a devil, myself. (I'm afraid I still believe entropy as a word humans use because they still haven't grasped the fullness of energy released nor where it really goes.)
To mg: I find your example very good. It has me mulling it over quite a bit. Does the algorithm meet my definition of thinking? I could say "yes," when considering the nature of my definition. But the other part of my definition is "thinking to survive." Can a computer "think" to survive on its own? As a species?

>>By redcatdave


I'll try to be as objective as possible:

Optimize: become closer to truth; involves learning.
(Computers: starting point IS truth, the one programmed)

Truth (absolute): reality, exists by it's own, has it's own rules; is random because of destiny? nature? god?
(Computers: humans programme computers to THEIR truth, limited, invariable once programmed)

"Truth" (relative): the one we create.
(Computers: their relative is the same as their absolute)

Experience: contact with reality (through perception). Builds up database of "truth". Induces optimization and allows to proyect expected results.
(Computers: They have their truth as a starting point. Can they increment their experience?)

So thought, for me, is sorting out not being restricted to programmed parameters (because our "truth" changes in the measure that through experience, we continously optimize it and have to validate it) because we can not be aware of reality (absolute) as somebody or something other from us, created it.

This is an attempt to be as objective as possible... It still can be optimized I guess. This is my "truth", the "truth" of a 23 year old Civil Engineer, which will be different to the "truth" of a 40 year old Civil Engineer or a 35 year old Artist.

>>By rafo


Kind of a fuzzy explosion. We all lift off, leaving every hard earned axiom behind. Landing somewhere in the unexplored land of natural language. It's raining little shiny questions. Truth? What is Truth? Definition? Absolute? Relative? To what? Reality? What's that? Somebody created it? How come we know that?

We probably reached the max, dancing in circles, discovering our own trails.

>>By mg


Are you fond on cars? Does anybody like talking about cars? Ferrari? 350HP? ABS Breaks? Maybe motorcycles? Harleys?

No, I'm joking. Yes, probably we reached the max. Maybe there's no way we can reach a definition that works for everybody. I was reading the other day that if we don't know which are the goals of humanity, how could we know that we are moving in the right direction? That we are being efficient? Maybe humanity is to become extint and therefore we are not helping much by thinking.

>>By rafo


Let's make a haiku...

To define a word,
You need clearly defined words.
Yikes! An endless loop.

>>By mg


Exactly, one cannot explain a language without using a language that must be explained with language. It is similar to Godel's idea that no mathematical system can be true and complete. No language can be true and complete.

>>By pgrmDave


I shall have to visit this discussion corner again . . . Reading "Godel Escher Bach" really was a dynamic experience for me . . . I was wondering what the group here might think of Richard Power's novel "Galatea 2.1"....I think there is a fascinating literary spin on AI there (which I first got seriously into via Hofstadter).....Powers is such a great writer in my book!
I have not read the full discussion here as of yet....but shall come back here at a more convenient time and peruse what's going on then.
Thanks for Your attention.....please note I am an amateur in these matters but they are quite fascinating (to say the least).

>>By satorotas


Hi Satorotas,

I haven't read Richard Powers' novel, "Galatea 2.1." But now you have me curious. I'll have to look into it!

>>By Cynth


I just read "I am a strange Loop".

In the past, when I read books by Hofstadter, each page was really intense. This time, on most of the pages I thought "ok, now get to the point, god damn it!".

However, it still was interesting and iaasl is an important book to me. Because it shows, how Hofstadter looks at the topic of consciousness these days (its been 20 or 30 years since he wrote Gödel, Escher, Bach I think?). He is much less interested in Zen these days. His main point now is, that the "I" (Or the "consciousness") is not a local thing. That "I" is nothing that can be located somewhere. And that our impression, that "I" is local is an illusion.

I like that thought and I like the image he uses: Cameras, that are pointed at the screen they are connected to. They see what they see. But they also see other screens connected to other cameras. These Camera/Screen combinations are our bodies. The "You" and "Me"s are not the bodies. The "You" and "Me"s are some structures created by these multidevice-feedback-loops and each "You" or "Me" structure is spread across multiple Screens.

>>By mg


Just flipped through the discussion quickly....The earlier question concerning what it is like to be a cat-and the question of the presence of thought and language. For an excellent essay on this topic, read Nagel. He attempts to do just that, but from the perspective of a hummingbird. But from our own observations, we can all probably infer some kind of thinking from basic behaviors. The question is, are we looking at the same kind of thinking as human thought, or just a varying degree? Does a cat have a sense of "self?" In the human brain/mind (which I certainly believe are on in the same) the illusive "self" is sometimes described as a representational system that contains not only an updating simulation of the external world, but which includes the simulating system within the simulation itself. It is self-referential, and as such, it updates the dynamic picture of the external, and its own involvement in it, constantly. Because we percieve causal connections everywhere, the brain also inputes causal functions to itself, as an active agent effecting exactly what it can simulate. Looping is what seems to be happening all the time.

I'd say that one of the best philosophers of science and mind today is Daniel Dennet, who co-wrote "The Mind's I" with Hofstadter. He works through the thought-language problems very concisely. One pervasive idea of his is that consciousness, as well as language, is evolving, and have evolved from less complex states. Countering the traditional dualism that seems to plague the infant mind, he argues like Hofstadter that the "I" is not an immaterial thing, but is a construct of a brain that is essentially billions of tiny mindless robot-like neurons. Consciousness arises, and can be described on several levels. From the physical stance, it is a massively parallel computer with billions of interconnected cells, using bio-chemical and electrical feeds to input, process, store, and output huge amount of information. But then there is the design stance, which is a level up, and helps to explain the brain from an evolutionary point of view. Why does this region exist, and how was it selected for. Next up is the intentional stance, which is where we almost always find ourselves. This perspective allows us to infer and act on the fact that other things in the world are also conscious. I do not come to the conclusion that you are an active agent in the world by measuring the activity of your neurons, and following the electrons bit by bit in your brain. I can even skip the design level and move up to the intentional stance, becoming instantly aware that if I want to survive is this dangerous world, my brain has no time to waste on such detailed investigation.

Just imagine a world in which every user had to know the phsyics of computers in order to operate them. Luckily, we can elevate our minds to the design stance and understand that this boxy thing has a function, a purpose, and proceed from there. But does a cat do the same thing? We cannot escape subjectivity, no, so there is an ultimate limit to where we can peer in nature. Is my green just like yours? Through objective means, we can only understand nature, and minds in particular, to an unsatisfying degree. One important question that biologists might ask is whether or not natural selection has taken felines down the path to self-consciousness, and if so, what were the pressures? Is there any evidence that cats have evolved higher order thinking? Does that kind of thinking only derive from social species? Is symbolic language, like our own, become a by-product of higher order thinking, or is it necessary for it?

Anyway, sorry about the rantish entry. I love the questions presented in your discussion, as well as the points made. Im always left thinking, which is probably a sign that this is the most important kind of discussion.

>>By Hume Ungus


You are not logged in. To access all functionality, you can log in via Flork.

Flork is a worldwide community of people who are interested in music, movies and books. As a Flork-User, you can participate in this and other discussions, meet new people and exchange messages with other members directly: www.flork.com