Intelligence defined.
by chunkdz
Recently, TT member "Thought Provoker" offered this very thought provoking comment.
"Practically every definition of "intelligence" includes referenced to the ability to learn, adapt, etc. An omniscient being isn't technically intelligent because he/she/it can’t learn anything new."
This brought to mind a cross-disciplinary generalized definition of intelligence that we here at TT brainstormed up some time ago.
Where:
PS = Problem Solving = An unweighted, finite, non-negative integer representing the number of problems a machine is able to solve.
P = Planning = A measure of a machine's accessible planning database.
L = Learning = Rate of successful inference from acquired sensory input.
IP = Intelligence Product = (PS)(P)(L)
Now let's plug in the numbers for an omniscient designer to see if Thought Provoker's assertion holds water.
Where:
PS = Infinite
P = Infinite
L = 0IP = Indeterminate
As we can see, even if learning is zero, the intelligence of an omniscient being is indeterminate. Indeterminate is far different from "isn't technically intelligent". It also is adequately descriptive of the only omniscient designer I happen to know of.
"Great is the LORD, and greatly to be praised; and his greatness is unsearchable." Ps 145:3
October 13th, 2010 at 11:56 pm
Then the Lord answered Job from the whirlwind: "Why are you using your ignorance to deny my providence?"- Job38:1-2
Comment by Bradford — October 13, 2010 @ 11:56 pm
October 14th, 2010 at 7:39 am
As Isaid in that thread thought provoker nes to think before he posts.
Not every definition of a word applies all the time.
Besides Dembski defined "intelligence" as it pertains to "Intelligent Design" years ago:
Intelligent Design is Nt Optimal Design
It's easy really- you cannot tel anything about the designer's ability to learn just by studying what it left behind.
You can't tell if the deigner has any problem-solving abilities just by studying the design.
And you really can't tell if the designer planned tat design- we can infer/ assume that bsed on the number of parts required- but that doesn't mean the designer actually did plan ahead.
intelligent
Comment by ID guy — October 14, 2010 @ 7:39 am
October 14th, 2010 at 10:03 am
Unless the Designer takes an IQ test and scores over 100, we can't be sure.
Arguably the Designer is the only "know-it-all" for whom that is a simple, non-pejorative, statement of fact.
But seriously, learning is not a requirement of intelligence.
Comment by pds — October 14, 2010 @ 10:03 am
October 14th, 2010 at 10:07 am
pds wrote:
I totally agree. Here is a nice proof: Intelligent Reasoning. The author of that blog is unable to learn even simple scientific concepts but he nonetheless claims to be intelligent.
Comment by olegt — October 14, 2010 @ 10:07 am
October 14th, 2010 at 10:09 am
chunkdz,
Are you assuming the Universe isn’t finite?
If the universe is finite, then PS is also finite.
As far a planning (P), I still have problems understanding how an omniscient being makes decisions.
How can he/she/it decide to do anything other than what is already known will be decided.
I have no problem with the suggestion your concept of God is beyond our ability to understand. He might even transcend logic itself.
At the very least, it looks like you agree it is “indeterminate” as to whether or not an omniscient Designer is technically intelligent. It would seem to me, it would be more universally agreed that whoever/whatever the Designer is, this entity would be purposeful.
Comment by Thought Provoker — October 14, 2010 @ 10:09 am
October 14th, 2010 at 10:39 am
Sorry Chunkdz,
I addressed my comment to Bradford instead of you.
The edit feature doesn't seem to work.
Comment by Thought Provoker — October 14, 2010 @ 10:39 am
October 14th, 2010 at 10:54 am
Fixed it TP.
Comment by Bradford — October 14, 2010 @ 10:54 am
October 14th, 2010 at 12:53 pm
pds:
IQ was a touch too anthropocentric. IP (Intelligence Product – see above) is much more universally applicable. I even ran the numbers for a tic-tac-toe machine just to give it a test run.
pds:
So you would agree that this Telic Thoughts definition of intelligence is far superior to what TP refers to as "practically every definition of intelligence"?
Comment by chunkdz — October 14, 2010 @ 12:53 pm
October 14th, 2010 at 1:12 pm
Hi TP:
No, only that the designer's knowledge is infinite.
"P" refers to problems that the machine can solve, not problems it has solved or problems it will solve. The ability is the important thing.
Granted.
If by "technically" you mean mathematically demonstrable, then yes.
But if we were to subtract just a single infinitessimal bit of knowledge from our omniscient designer's mind, it would nearly approach infinite intelligence.
We can therefore reasonably justify the belief that an omniscient designer is, at least, infinitely intelligent. It just can't be demonstrated it mathematically, afaik.
Not sure how you arrive at this conclusion. On it's face I would disagree.
TP, would you agree that the above definition of intelligence is far superior to what you refer to as "practically every definition of intelligence"?
Comment by chunkdz — October 14, 2010 @ 1:12 pm
October 14th, 2010 at 1:31 pm
Hiding behind definitions in order to dismiss conclusions is such a petty debate tactic.
Even at face value this statement is false. Having the ability to learn does not require having available something to learn so omniscience is not logically contradictory of having an ability to learn.
There are countless things I have the ability to accomplish even though I have no opportunity to do so … and even though there may never be such an opportunity, nor even could be.
I have the ability to digest jelly beans although there are none available. I have the ability to digest jelly beans made in the shape of Man O War, though none exist.
I have the ability to drive a car 1,000 miles per hour. I have the ability to live on an Earth-like planet in a distance galaxy.
Comment by Pez — October 14, 2010 @ 1:31 pm
October 14th, 2010 at 1:51 pm
Pez,
You wrote…
It wasn't debating. I was answering pds's question "What do you think?"
Please note I tried to avoid this subject in my original answer.
I made a mistake answering this thread.
I thought Bradford had started it. Had I noticed it was Chunkdz, I probably wouldn't have participated in it at all.
Comment by Thought Provoker — October 14, 2010 @ 1:51 pm
October 14th, 2010 at 2:04 pm
I'm being nice, TP, and I sincerely thought your comment was genuinely provocative. And I hope you find my response just as engaging.
Life's too short to hold grudges. C'mon, let's do science!
Comment by chunkdz — October 14, 2010 @ 2:04 pm
October 14th, 2010 at 3:12 pm
Yes, I agree. But I don't think you need that to conclude that an omniscient being is intelligent. I think it is obvious, which is why only the last sentence of my comment was serious.
Comment by pds — October 14, 2010 @ 3:12 pm
October 14th, 2010 at 3:32 pm
Hi chunkdz,
ok, I'll try to play nice…
I found your tic-tac-toe example (link)
My versions…
We'll measure the intelligence of a universe than exists.
PS = 1, problems solvable by universe
P = 2, all possible outcomes (exist or not exist)
L = 1 possible successful strategy from 2 choices
Where I = (PS)(P)(L) = 2
———————————————————————Can we calculate the intelligence of Earth's biomass?
PS = ? problems solvable by biomass (existance, growth, interaction, etc)
P = ? all possible outcomes (fossil evidence suggests quite a lot)
L = ? possible successful strategies from inputs (at least 1)
Where I = (PS)(P)(L) = ? (greater than 0)
I suggest that simply using your equation it can be argued Mainstream Evolutionary Theory (i.e. "Darwinism") is an example of intelligent design.
Comment by Thought Provoker — October 14, 2010 @ 3:32 pm
October 14th, 2010 at 4:20 pm
TP,
I think you have a misconception about (P). It isn't simply a tally of all possible outcomes. It is the size, in bits, of a machine's accessible planning database.
In the Tic-Tac-Toe example, the machine has access to a database representing all possible states so that it can plan it's next move.
In regards to Darwinism, life is not supposed to be able to plan ahead – much less have access to a database of possible future states. (However if you were to demonstrate that life DOES plan ahead by choosing future states from a database this would support an intelligent design inference.)
Also in regards to (PS): Problem solving is simply the ability to transition from one state to another. For the universe to solve the 'problem' of existing or not existing it would need to be able to transition from a state of existing to a state of not existing and back again. Otherwise (PS)=0.
Comment by chunkdz — October 14, 2010 @ 4:20 pm
October 14th, 2010 at 5:43 pm
Hi Chunkdz,
You wrote…
I have got to run. Please read this old thread…
An Atheist's view of ID, Front Loading and Retrocausality
I don't know if you remember it, but you made a comment in it.
Kornbelt made a lot of comments as well as Bradford.
until later…
Comment by Thought Provoker — October 14, 2010 @ 5:43 pm
October 14th, 2010 at 6:40 pm
TP:
I remember.
Like I said…
Perhaps you could do another guest post and update us on what has happened in the world of Orch O/R over the past 3 years.
Comment by chunkdz — October 14, 2010 @ 6:40 pm
October 14th, 2010 at 9:25 pm
Look "Intelligent Design" just refers to agency involvement.
counterflow
It is all about what is required to bring the result observed.
Was agency involvement required?
Or, as Del puts it, can nature, operating freely, account for it?
Again all the power is in their hands- just show that blind, undirecetd processes can account for it and according to Del, Dembski, Meyer, Behe- well all IDists- there isn't any requirement for agency involvement and therefore no design inference.
But what do they do with that power?
Well look at olegt- all that scholarship and all he has are school-yard taunts.
Is it just me or does anyone else get the feeling that evolutionists were the kids no one picked to be on their team in school?
Comment by ID guy — October 14, 2010 @ 9:25 pm
October 14th, 2010 at 10:02 pm
Hi Chunkdz,
You suggested…
Before I commented here a couple of days ago I nosed around a little and found some recent papers on Quantum Biophysics. While things are progressing pretty much as I expected, it is mostly confirmation biological quantum effects do happen.
However, your offer compelled me to look something more worthy of a guest post.
I found a paper that is… err… umm… interesting. Hopefully, thought provoking.
this
Its tone and subject material sound like something bright 8th graders would write, especially in the beginning.
But if you push through and read the whole thing, I don't see anything clearly wrong with it, other than it invokes a strong feeling of incredulity.
I will write something up this evening to present and defend it.
if you could start a thread titled
Emergence of the physical world from information processing
and include an introduction and this image…
http://4.bp.blogspot.com/_tua8...
I will post my comment.
Thanks
Comment by Thought Provoker — October 14, 2010 @ 10:02 pm
October 15th, 2010 at 12:05 am
I'm not a materialist. I am comfortable with Penrose’s version of the Copenhagen interpretation and the “reality” of wavefunctions which are no more tangible than radio waves. However, I still think of nature as having some kind of objective reality. Thus making me a dreaded "naturalist".
Therefore, this paper has the potential of making me rethink my worldview.
I am curious as to how others will receive it.
Its premise is fairly simple, almost too simple.
Historically, philosophers have struggled with the objective physical reality verses the ideal spiritual reality. From the paper…
Could it be there really is only one truth and I have to give up my NOMA security blanket?!?!?
At this point, this is about as Worldview challenging as watching the Matrix movie. But let's hear the authors out. Look at this picture…
http://4.bp.blogspot.com/_tua8...
The paper goes into a lot more detail, but just of it is that there are three choices.
A- We are objectively real observing an objectively real world.
B- We are objectively real observing a virtual world.
C- We are virtual observing a virtual world.
The paper is hypothesizing “C” is correct.
The big hump to overcome the extremely strong presumption that an objective reality must exist, at least somewhere. For example, there must be at least some kind of computer running a simulation and it is the computer that is objectively real.
But there is more. From the paper…
The authors go into more depth and I would encourage people to read the paper for themselves. To help the reader organize a response they offered a multiple choice summary…
I am especially interested in hearing what Oleg has to say.
Comment by Thought Provoker — October 15, 2010 @ 12:05 am
October 15th, 2010 at 12:09 am
Chunkdz,
If you would, please replace the link to the picture with the image and call the thread.
Emergence of the physical world from information processing
Comment by Thought Provoker — October 15, 2010 @ 12:09 am
October 15th, 2010 at 4:10 am
Interesting, TP.
Comment by computerist — October 15, 2010 @ 4:10 am
October 15th, 2010 at 9:58 am
Interesting indeed. I have no difficulty accepting the concept that the physical world could be an outgrowth of information processing and if ID were described in such terms that too would be fine with me. Whether the process and the entity generating the process is traced to God or not can be left to a theological discussion.
My theology is that faith in God is enabled by God and only through a free will acceptance of that divine initiative. So the idea that ID equates to a belief in God or particularly the God described in the NT, is a false one. That in turn means the stalking horse concept is a theological non-starter.
Comment by Bradford — October 15, 2010 @ 9:58 am
October 15th, 2010 at 3:49 pm
No problem. But first, answer my previous question.
I'll get to the new post this afternoon. Crazy busy.
Comment by chunkdz — October 15, 2010 @ 3:49 pm
October 15th, 2010 at 5:10 pm
Strangely enough, I remember reading somewhere that there did exist a superluminal photon. http://cerncourier.com/cws/art...
Gravity is a bizarre component of our existence. According to the paper that TP lists, this tidbit:
Seems the superluminal photon hypothesized around curved-spacetimes might make this a difficult hurdle.
Comment by ciphertext — October 15, 2010 @ 5:10 pm
October 15th, 2010 at 5:16 pm
As interesting as I find simulation hypotheses, that paper has bugged me for a while. First because he equivocates mightily on what sort of "big bang" is "necessary" given a virtual reality. Second, all that pixel talk just seems so off-base. Why he thinks a computer monitor has much to do with a simulation eludes me, but it could be that he's just being sloppy here.
Comment by nullasalus — October 15, 2010 @ 5:16 pm
October 16th, 2010 at 1:57 pm
Chunkdz,
You asked…
The reason I brought up the Retrocausality thread was to point out your simple equation probably doesn't fully explain anyone's idea of what "intelligence" means, even yours.
A simple tic-tac-toe program "predicts" the future via a database of all possible outcomes.
There is also the problem of deciding what is and isn't a "problem". (pardon the pun).
A simple electronic latch circuit solves the "problem" as to whether a light should turn on or off when a button is pressed. Is it intelligent?
A simple tic-tac-toe machine is basically a latching circuit with nine buttons and nine bicolored lights. With the right incentive (i.e. getting paid for it) I could build such a machine.
And finally there is the definition of "successful strategy".
Who/what gets to decide what is successful outcome?
A "successful" tic-tac-toe machine could be one that challenges a toddler but lets him/her win, thus improving self image.
A "successful" tic-tac-toe machine could be one that never wins or loses no matter what the opponent does.
Let me stop there in my attempt to "play nice".
Comment by Thought Provoker — October 16, 2010 @ 1:57 pm
October 16th, 2010 at 2:21 pm
TP:
Not true. Plug in some numbers from one of your examples of retrocausality and find out for yourself.
I've already defined this. Problem solving is the ability to transition from one state to another. Therefore a problem is simply that – a state.
Plug in the numbers and find out.
It's "successful inference". And success can be defined either by the designer, or by a curious onlooker using critical thinking.
Yup. Does it bother you that a machine might not meet your expectations? Does that mean we cannot measure intelligence without independent knowledge of the designer.
Not sure what you mean. I thought we were 'doing science'.
It would help, however, if you would simply admit that the above equation is far superior to practically every definition of intelligence you've seen – if only by merit of not possessing the fatal flaw which you referred to.
Comment by chunkdz — October 16, 2010 @ 2:21 pm
October 16th, 2010 at 4:33 pm
Chunkdz,
"play nice" and "doing science" can sometimes conflict.
This is my third comment where I am telling you where I disagree with your conclusions (doing science).
PS = number of state transition sequences
P = LOGn (number of possible sequences) = database of Good/Bad decisions
L = LOGn (Number of successful outcomes /number of inputs)
Your tic-tac-toe example…
We'll measure the intelligence of a machine that plays Tic-Tac-Toe.
PS = 362,880 (9!)
P = 20 (18?)
L = 2 (8 / 4)
IP = 362,880 * 18 * 2 = 13,063,680
I'm not sure how you calculate 20 or 18 for P or how you decided there were 8 "successful strategies". I suspect they are somehow related.
Here is my version…
PS = 255,168 (possible games are less than 9! because winning stops the sequences)
P = LOGn(255,168) = 12.45
L = LOGn(131,184/9) = 9.59 bits
IP = 255,168 * 12.45 * 9.59 = 3.0 x 10^7
If you define "successful strategy" as winning whenever possible when the tic-tac-toe machine has the first move (i.e. is "X").
This calculation makes no assumption about implimentation. A 9.59 bit database would imply a tic-tac-toe would need at least a 97 bytes of storage. I am pretty sure I could right a tic-tac-toe routine using less than 97 bytes for both database and logic.
I will stop here to give you a chance to further explain how you arrived at your numbers for your tic-tac-toe example.
Comment by Thought Provoker — October 16, 2010 @ 4:33 pm
October 16th, 2010 at 4:36 pm
oops, I used 9.59 bits when talking about writing a routine. I should of used 12.45 bits. This strengthens the point I was trying to make.
Comment by Thought Provoker — October 16, 2010 @ 4:36 pm
October 16th, 2010 at 9:48 pm
Even if my math were wrong the formula is still valid.
Even if one tic-tac-toe machine were more intelligent or less intelligent than another tic-tac-toe, the formula still works.
Which is…?
Comment by chunkdz — October 16, 2010 @ 9:48 pm
October 16th, 2010 at 10:36 pm
Chunkdz,
I am trying to make sure I am applying your formula correctly for the most detailed example you have give, the tic-tac-toe machine.
Let me do it again so you can tell me if and where I am applying it incorrectly.
PS = number of state transition sequences
P = LOGn (number of possible sequences)
L = LOGn (Number of successful outcomes /number of inputs)
PS = 255,168 which is all possible valid tic-tac-toe games (unique state transitions from start to finish).
P = LOGn (PS) = 12.45
This represents the "database" of Good/Bad indicators. One bit per unique sequence is needed. I am not sure why both PS and P are needed. It would seem that P alone would be good enough.
While the parameter "P" makes some sense to me, the parameter "L" is mostly a mystery. I am not going to try guessing anymore.
Of the possible 255,168 games player "X" Wins 131,184, Loses 77,904 and Draws 46,080
Would you please detail how you calculate parameter "L"?
Do you agree with how I calculated "PS" and "P"?
Comment by Thought Provoker — October 16, 2010 @ 10:36 pm
October 17th, 2010 at 12:34 am
TP, we're not trying to design DEEP BLUE for Tic-Tac-Toe here.
I just made an example to demonstrate the power of the IP formula.
The numbers could change depending on the goals of the machine, or whether it always starts with X or O, or if it plays to tie, or plays to win, or plays to lose even. It seems you've got the gist of it except for (L). (L) is simply the number of successful inferences via the sensory inputs. The example I used previously was with a DARPA car. A car that could successfully recognize a buffalo in the road via it's on board camera (sensory input) would have a higher L value than a car that couldn't make the successful inference.
Now…since my formula was able to absorb your most potent criticism of "practically every definition of intelligence", wouldn't you agree that my definition of intelligence is far superior to practically every definition of intelligence?
Comment by chunkdz — October 17, 2010 @ 12:34 am
October 17th, 2010 at 12:47 am
Is a push button lamp switch intellegent?
PS = number of state transition sequences = 2 (ON to OFF, ON to OFF)
P = LOGn (number of possible sequences) = LOGn (2) = 1
L = LOGn (2/1) = 1
IP = 2*1*1 = 2
If you are comfortable with this, I guess I can accept it.
But, no, I do not think it is a very useful metric.
Comment by Thought Provoker — October 17, 2010 @ 12:47 am
October 17th, 2010 at 10:18 am
TP
Please stop redefining the parameters. If you read the OP you will find the actual definition of PS:
A push button lamp switch does not have the ability to transition states without outside intelligent intervention. PS=0.
Put in a light sensor, some logic circuitry and some transistors and you may be able to bump up that PS value.
Like my Dad used to say: "If all else fails, read the instruction manual".
Comment by chunkdz — October 17, 2010 @ 10:18 am
October 17th, 2010 at 10:19 am
Comment by chunkdz — October 17, 2010 @ 10:19 am
October 17th, 2010 at 10:52 am
A push button lamp switch needs an intelligent agency in order to come into existence.
So a push button lamp switch is evidence for intelligence.
Comment by ID guy — October 17, 2010 @ 10:52 am
October 17th, 2010 at 1:22 pm
Hi Chunkdz,
Earlier you wrote…
A toggle switch has "…the ability to transition from one state to another" and back again.
The reason I feel your formula is not a very useful metric is because either it is simple enough to include a toggle switch or there are hidden subjective inputs or both. At the very least you seem to agree the parameter "L" is inherently subjective.
I am comfortable with the amount of effort I expended in trying to understand it.
I have no interest in continuing this conversation.
Comment by Thought Provoker — October 17, 2010 @ 1:22 pm
October 17th, 2010 at 4:42 pm
It's funny to me that this argument is even happening.
The point of it all is to come to some agreement as to whether life required intelligent intervention or not.
The funny thing (again – to me) is this: If you accept that life was intelligently designed and then examine the level of intelligence that would be required to design life, you fairly quickly arrive at the conclusion that only an intelligence capable of knowing all possible outcomes of all possible chemical combinations in all possible environments (i.e. omniscient) could have designed life.
If you think I'm overstating, read this and ruminate on it for awhile while asking yourself, "what level of intelligence would be required to figure out how to fit all of this together?"
Comment by Daniel Smith — October 17, 2010 @ 4:42 pm
October 18th, 2010 at 10:15 am
TP:
Not without intervention.
Perhaps the definition wasn't clear enough. I think I'll add the word "autonomously" from now on. I appreciate the input.
The toggle switch confusion was simply a misunderstanding. As for inherently subjective – once the goal is set by the designer the "L" parameter is no longer subjective. In other words, once the designer decides that the goal is win at tic-tac-toe, then there are eight winning strategies for the machine to learn, and L becomes a known quantity.
I suspect you are letting your personal feelings get in the way of critical thinking. There's nothing really very controversial about saying that intelligence is the product of problem solving, planning and learning. Don't let your feelings of animosity towards me blind you to the incredible power of the Intelligence Product formula. I think you'll agree that it defines intelligence more effectively than practically every definition of intelligence out there.
Comment by chunkdz — October 18, 2010 @ 10:15 am
October 18th, 2010 at 11:26 am
Let's clarify the toggle switch example which TP brought up.
A toggle switch on it's own is not able to solve a problem (transition). (Unless you want to call slowly rusting and decaying a "solution" to a problem)
But how about adding an IR motion sensor?
PS=2 since there are now two problems which the machine is able to solve autonomously. (1) Turn on when someone walks into the room, and (2) turn off after a pre-defined period of inactivity.
P=1. P is simply a machine's "road map". 2 states are representable by a single bit.
L=1. There is a single successful inference that this machine can make – a person is in the room. Assuming there is only one sensor, the input is 1 bit as well. Since most consumer motion detectors must have their threshold manually adjusted, they are not really learning. But let's assume this one has some kind of internal feedback system which allows it to self adjust it's threshold and learn what makes a successful inference.
IP=(2)*(1)*(1) = 2
Not very intelligent compared to our tic-tac-toe machine which had an IP in the millions, but still demonstrating the most rudimentary artificial intelligence.
Once again, the IP proves that it is the most effective definition of intelligence over practically every other definition.
Comment by chunkdz — October 18, 2010 @ 11:26 am
October 18th, 2010 at 11:46 am
Daniel Smith:
First things first. One of the most common objections from ID critics is the fact that Intelligence is often poorly or loosely defined. I'm sure you've experienced it yourself on occasion – you use the word 'intelligence' in an argument and some critic will inevitably demand a definition, and subsequently complain that your definition is inadequate.
Well no more. Now that Telic Thoughts has rigorously defined intelligence in a general and cross-disciplinary way, you can silence the critics immediately and move on to the more substantive arguments.
Critic: "Hey you IDiot, you haven't even defined intelligence so shut up."
You: "You couldn't be more wrong! Intelligence is defined as:
Comment by chunkdz — October 18, 2010 @ 11:46 am
October 18th, 2010 at 5:47 pm
Null, he's merely using the pixel analogy to convey the idea of how algorithms within the "computer" that is "implementing the simulation" could be the source of the hidden instantaneous non-local transfer of information that have been experimentally verified within spacetime as we know it. And also to demonstrate how a subatomic particle could hop from "here" to "there" without traversing any space between. It happens because there is no space to traverse, per se, just like when "moving" a pixel across a video screen. It's an immediate change of state from one position to another in a informational matrix.
The bottom line is, the algorithmic, computation view does explain what's going on. Whether it's the only view that can, or the best view, is another matter. Of course, it doesn't tell us the nature of the "computation device."
Comment by kornbelt888 — October 18, 2010 @ 5:47 pm
October 18th, 2010 at 7:19 pm
Typical. Promise a definition of "Intelligence", then, instead, provide a definition for "Intelligence Product". Even if the definition were useful, it's still not a definition of "Intelligence" until you actually put a stake in the ground and say "An Intelligence Product greater than N means intelligence is present." Can you do that? If so, can you run through an example, please? Are porpoise intelligent, for example?
Comment by don provan — October 18, 2010 @ 7:19 pm
October 18th, 2010 at 7:34 pm
Can I try?
Intelligence = having a mind and using it
Comment by Daniel Smith — October 18, 2010 @ 7:34 pm
October 18th, 2010 at 10:42 pm
Intelligence: the ability to run scenarios about the future based on a current database of information ("foresight"), ability to make choices, ability to learn new facts, having goals. Isn't this the essence of intelligence? Why is this so difficult?
Comment by kornbelt888 — October 18, 2010 @ 10:42 pm
October 18th, 2010 at 11:04 pm
Hi Kornbelt,
I think your definition for "intelligence" is well reasoned and reasonable.
Comment by Thought Provoker — October 18, 2010 @ 11:04 pm
October 19th, 2010 at 12:21 am
kornbelt:
Kornbelt, by your definition an omniscient being isn't technically intelligent because he/she/it can’t learn anything new. This is the fatal flaw that Thought Provoker pointed out in the OP.
Comment by chunkdz — October 19, 2010 @ 12:21 am
October 19th, 2010 at 12:24 am
TP:
It is fascinating how a person's emotions can cloud rational thought.
Comment by chunkdz — October 19, 2010 @ 12:24 am
October 19th, 2010 at 12:27 am
Provan:
Any positive value for IP = intelligence. If you read the thread you'll find several examples from light switches to tic-tac-toe machines.
Comment by chunkdz — October 19, 2010 @ 12:27 am
October 19th, 2010 at 7:42 am
Daniel Smiith
Ok, but remember what you said earlier:
It's not enough to simply declare a mind intelligent. As you said, we want to distinguish levels of intelligence. We need a method to measure intelligence that is not anthropocentric.
We now have that method. It is called The Intelligence Product.
Comment by chunkdz — October 19, 2010 @ 7:42 am
October 19th, 2010 at 10:37 am
But TP's so called fatal flaw is fatally flawed because ti is motivated by his desire to avoid an omniscient designer. The ability to learn is not dependent upon there being something left to learn. Certainly omniscience includes the knowledge of how to assimilate new information, were there to be any.
Comment by Pez — October 19, 2010 @ 10:37 am
October 19th, 2010 at 11:22 am
Oh, yes, sorry. I'm just now catching up. So the IR sensor has an IP of 2, the tic-tac-toe machine in the millions, so both are intelligent according to your definition. That doesn't seem very useful to me. Specifically it seems obvious that the process biologists call "evolution" would be intelligent according to this definition.
In this context, I think "technically" means "according to your definition".
You want to change the definition of God to remove the quality of "omniscient" and replace it with "nearly omniscient"?
We're not talking about reasonable justification. We're talking about a definition of intelligence. Your definition cannot tell us whether God is intelligent.
This really illuminates the core issue of the whole "define intelligence" objection. You've provided a definition, and you've agreed that the definition does not identify God as intelligent. Yet we all agree that God could be the designer, if He existed. Therefore we must conclude that the design of what we see around us might not require intelligence as you have defined it.
Do you see? You've carefully provided a specific concrete definition, problematic as it is, but then you have to abandon that very definition and fall back to "reasonably justify" to support the ID conclusion.
Comment by don provan — October 19, 2010 @ 11:22 am
October 19th, 2010 at 1:48 pm
Provan:
It is useful for distinguishing that which is intelligent from that which isn't. It is also useful for distinguishing that which has higher intelligence from that which has lower intelligence.
If you had read the thread as I suggested you would have noticed that TP already raised this objection, and the objection was overruled by the fact that "evolution" does not plan for the future. P = 0, IP = 0.
If you had read the thread, you would have noticed that TP said "technically" before I even wrote the OP.
Had you actually read what I said you would have realized that I was not suggesting we change the definition of God.
Yes it can. But it requires analysis.
Now we're getting somewhere.
So let's go back to the tic-tac-toe machine. It's (L) value is directly related to there being 8 successful win strategies to learn via it's 9 sensory inputs.
Once the machine has learned every strategy it is no longer capable of learning anything new. Does that mean that it is no longer intelligent? Of course not.
Now apply this logic to an omniscient being. "All-knowing" means that He has already learned everything there is to learn. But to determine (L) we need to know something about our omniscient being. First we need to know the number of things there are to learn. This would be infinite. Then we need to divide that number by the number of bits of sensory input that the omniscient designer has. This is either infinite or indeterminate, but let's keep the argument going by suggesting it is infinite.
So (L) for an omniscient being could then be expressed as infinity/infinity which is indeterminate.
Summing up: whether we express (L) as zero, or as indeterminate, we still get the answer IP = indeterminate.
No. That's like saying we have to abandon the Theory of Relativity to reasonably say that the center of a black hole has mass.
We don't abandon mathematics simply because certain answers are indeterminate. We simply apply mathematical analysis. Calculus is such an analytical methodology built around the idea of infinitessimally small, indeterminate objects called limits. Infinity is simply an unbounded upper limit.
Comment by chunkdz — October 19, 2010 @ 1:48 pm
October 19th, 2010 at 2:26 pm
Also, everyone should take a gander at this classic example of disconfirmation bias in action.
Don Provan on the evening of the 18th.
Chunkdz, later that evening.
Provan, the next morning:
In the mind of this type of ID critic, anything that might support an ID argument is by default dishonest, useless, wrong, and vacuuous. Don amply demonstrates this by coming to these conclusions more than fifteen hours before reading the thread.
As always, folks, it is important to recognize that these are not rational people we are dealing with. Always keep this in mind.
Fascinating.
Comment by chunkdz — October 19, 2010 @ 2:26 pm
October 19th, 2010 at 2:47 pm
The point is that a test which identifies an IR sensor as intelligent is not very useful.
What part of a general purpose immune system is not "plan for the future"? Evolution plans for the future just like you do: it tries various things, discovers which work well, and continues to use them so that they are available in the future.
I think I just caught out of the corner of my eye that you're complaining I'm not paying 100% attention to your thread, and I do apologize for that, but these fundamental points seem to me to require clarification before I can really understand your more detailed analysis, so I'm going to raise them even though I haven't had a chance to read over everything else here.
Comment by don provan — October 19, 2010 @ 2:47 pm
October 19th, 2010 at 5:03 pm
Provan:
If only you had bothered to read the post you would have noticed that I never identified an IR sensor as intelligent. Yet here you are, sneering and poo-pooing like an ID critic in full blown disconfirmation bias – having never bothered to read that which you are criticising.
It is fascinating.
Actually when I plan a trip, I find the address where I want to go, plan out the best route to take, and drive there. I don't "try various things".
You've made another common scientistic mistake of equating life with evolution. Life might be described as intelligent, but life is not evolution, and evolution is not life.
Not complaining. Fascinated.
For most people, criticizing something you haven't bothered to read and getting called on it elicits embarrassment. But in your case, you simply double-down and re-assert your decision to criticize that which you haven't read.
Actually I'd prefer that you read first, then ask questions. Do you think that's too much to ask?
Comment by chunkdz — October 19, 2010 @ 5:03 pm
October 19th, 2010 at 7:37 pm
Gotcha.
I was attempting to limit the definition to beings – just to avoid the common pitfalls we're already seeing here with the arguments about light switches being intelligent and so on. We generally can rate the intelligence of other beings without formulas and such. Of course that's all intuitive and non-sciency so I'll leave the scientific pursuit to you.
Comment by Daniel Smith — October 19, 2010 @ 7:37 pm
October 19th, 2010 at 8:01 pm
Daniel Smith:
When the common pitfall is willful ignorance there is little hope of that.
Comment by chunkdz — October 19, 2010 @ 8:01 pm
October 19th, 2010 at 8:27 pm
I didn't have an omni in mind with my definition. I suppose we could amend it and say an intelligence is something that has these powers or greater as exhibited in spacetime. A man can build a ship within spacetime. An omni can build a ship within spacetime, presumably. It does no violence to the intent to say that "intelligence" is working in both, IMO. At any rate, spacetime manifestations of intelligence are all we can imagine anyway. I see no point in bringing an omni into the discussion, quite frankly, because whatever such a thing might be, to us its products would bare the marks of something that was in possession of at least intelligence as I have defined it. Moreover, I have no evidence that an omni actually exists, and I have plently of evidence that non-omni intelligences exist.
Comment by kornbelt888 — October 19, 2010 @ 8:27 pm
October 19th, 2010 at 9:58 pm
Kornbelt:
Well put. In truth our definitions are not that different.
But isn't it nice to know that we now have a definition of intelligence that:
A) Can distinguish Intelligence from non-intelligence
B) Can distinguish different levels of intelligence from each other, and…
C) Can adequately and uniformly describe any intelligent machine from a basic AI switch to an omniscient being.
The Intelligence Product.
Embrace The Understated Elegance. Embrace The Power.
Comment by chunkdz — October 19, 2010 @ 9:58 pm
October 20th, 2010 at 2:26 am
HOO RAH
Comment by Guts — October 20, 2010 @ 2:26 am