A definition of Personality

[Luke Breuer]

I have a peculiar definition of personality, which, if valid, goes some very interesting places. At first, it may appear algorithmic and mechanical, but that is only because most people are not used to thinking of infinities in the way that I use them below. I first posted a version of the following on the website “Something Awful”, but the below did not pique their curiosity. I hope that I find something different here, what with the focus on infinities and absolutes. :-)

Personality: a definition

1. a formal system with infinitely many non-recursively enumerable axioms
2. a formal system with infinite [and uncompressible] definition

The mathematicians/CS majors will perhaps be satisfied with definition #1 and see a connection to Kurt Gödel’s incompleteness theorems. It may help others to realize that definition #2 is equivalent to #1: “recursively enumerable”, or “effective generation”, simply mean that the axioms cannot be generated by a computer with source code of finite length. (Strictly speaking: a Turing machine with finite starting tape.) Note that “axioms” can be replaced with “equations”, “basis vectors (QM)”, or even simply “states”.

An attempt at a less precise, layman’s version: you realize a person is a person and not a machine when he* never fails to keep on surprising you. You think you understand all of his little habits and quirks and then he does something unexpected. You can think of your continual learning about him as adding more and more axioms to your formal system describing him. Some of the learning can also involve correcting axioms and/or increasing their precision. The fact that you can never come up with a complete set of axioms indicates, to me, that there are infinitely many. Yes, this is a matter of faith, as it is based on inductive logic.

Interesting place #1: emotions
As long as we have not sufficiently characterized any “system”, it will appear to have personality, as defined above. I claim that emotions can be viewed as interference patterns (think of any kind of waves) between a mind and a system which appears to have personality; whether or not the system really is another mind—with “true” personality—is irrelevant. Not only does this explain the tendency to anthropomorphize, but it also provides an opportunity to add some precision to a topic that is often quite fuzzy. Examples:

• surprise when we find out our characterization wasn’t quite right
• frustration when we can’t improve our characterization; perhaps we find we’ve started to be consistently wrong
• anger when we “know” the system must work in a way that defies what we currently observe
• happiness when we learn more about how the system works, or perhaps when our model has been confirmed
• humor when we made a dumb assumption and it shows up so that we obviously see how we were wrong and what the right answer is
• love when the positive emotions outweigh the negative ones, on average (perhaps with some weighting of bad vs. good)
• hate when the negative emotions outweigh the positive ones

Remember that the above apply to anything that the observer can view as having personality, whether the full definition of the system is infinite or finite.

The above provides a view of emotions that does not necessarily question their usefulness (or cost/benefit ratio). Things become intriguing if you view emotions as an interference pattern between a subset of your personality you have not yet fully characterized and a subset of whatever you are observing. There are two ways to “better” a RE formal system: elimination of contradictions (inconsistencies) by altering/removing axioms, and elimination of Gödel sentences by altering/adding axioms. The former can use pure deduction and be emotion-free, but the latter seems more inductive and in need of some efficient mechanism to choose likely solutions in the “logical possibility space”. If indeed emotions can be viewed as interference patterns, they could be absolutely critical to this role. Then again, perhaps there is non-deductive “intuition” that can serve this role, which is distinct from “emotion”. I’ve hit the end of my comfort zone in terms of speculation here; feedback would be much appreciated.

Place #2: reality
Some scientists think that one day, they will find mathematics to perfectly describe reality. They may discover that hard determinism is true. That is, however, unproven—it can only be assumed, and perhaps assumed only poorly. If one were to use induction with base cases built upon history, one is forced to say that how much ever science is done—no matter how many questions are answered to increasing precision—even more mysteries and questions seem to open up. There does not seem to be an end in sight for mathematics, nor any other field. When Darwin wrote The Origin of Species, he thought the cell was a blob of goo. Now, while we know an incredible amount about the cell, our limited ideas of what is going on get disproved time and again. Yes, we come up with good models, but we keep finding out that there is even more we have yet to explore!

If we use proper induction—of the kind that results in the belief in naturalistic historicism**—we are forced to assume that there is no end in sight of how much science can be done. The result of that is that reality has infinite description, which means: reality has personality. This is a surprising conclusion, but I have yet to successfully poke any holes in it, or in my definition of personality. Perhaps someone else can. :-p

Place #3: other minds (or personalities)
I’m not sure what the status of the “Problem of Other Minds” is on this forum, nor the status of “Unity and Diversity”, but the above definition of “personality” seems to contribute to both ancient philosophical problems. First, if there are multiple minds, then their descriptions are not identical. Second, their descriptions can overlap, either approximately or perfectly (that is, in certain finite domains). Finally, the sum/union of all minds could be called “God”, or at least a subset of “God”. Within this model, personalities can interfere constructively and destructively. One can view each “axiom” of a personality to have a given amplitude; interaction between personalities can involve the increasing and decreasing of individual amplitudes.

One of the most famous passages in the Bible, Matthew 7:1-5, can be given a mechanism within this framework. When we interact with other personalities, we [hopefully] attempt to simulate them in our minds. Now, any simulation is necessarily flawed unless it is a perfect representation; since personalities have infinite definition, the only way for one person to perfectly simulate the other would require either A) the simulation to only require finite description, or B) the simulator to entirely contain the personality being simulated. If we note that situation A does not actually occur that often (or that we rarely get even finite descriptions perfect), then any simulation will involve “crack-filler”, to substitute for incomplete information. An all-too-common source of crack-filler is the simulator’s personality himself. Now, if I criticizing a simulation of even a subset of another mind, how do I know whether I am criticizing the valid part of the simulation and not invalid crack-filler? Indeed, perhaps the very nature of simulating builds upon the description of my personality, so that any intersection between my personality and the other personality tends to result in improperly increased or decreased amplitude!

Place #4: your thoughts?
If you find my definition of personality motivating, perhaps you can find other areas of knowledge/wisdom it touches upon?


* I use “he/she” when it is not too much effort, but if too many slashes are used due to repeated use of the pronoun, I often switch to exclusively-masculine for the sake of clarity.
** naturalistic historicism: “what usually happens is what always happens”

(Addition of [super][/super] and [sub][/sub] BBCode would be quite helpful.)

[cousinbasil]

An attempt at a less precise, layman’s version: you realize a person is a person and not a machine when he* never fails to keep on surprising you.

He would have to fail sometimes or else there would never be "surprises." It would have to be a matter of timing. Otherwise, you would "expect the unexpected" from such a person.

You think you understand all of his little habits and quirks and then he does something unexpected.

You can think of your continual learning about him as adding more and more axioms to your formal system describing him. Some of the learning can also involve correcting axioms and/or increasing their precision. The fact that you can never come up with a complete set of axioms indicates, to me, that there are infinitely many. Yes, this is a matter of faith, as it is based on inductive logic.

You are saying what it would require for one to be able to discern between a person and a machine. A characteristic of one person versus another person are just those behaviors that are repetitive and predictable.

An entity that always surprised you and was completely unpredictable would not resemble a real person at all.

[cousinbasil]

As far as the Turing test goes, you have to take into account which human being your AI device is "attempting" to fool into believing he is communicating with another person. For instance, AI was advanced enough as early as 1982 to trick David Hasselhoff into having a personal relationship with his car K.I.T.T. I understand this required almost two dozen lines of code.

[Luke Breuer]

He would have to fail sometimes or else there would never be "surprises." It would have to be a matter of timing. Otherwise, you would "expect the unexpected" from such a person.

I was looking for a way to say that there was not a time T, past which the person never surprised you again. I’m not entirely sure how valid the phrase “expect the unexpected” is—how often does it mean that either A) the person is irrational, or B) that you really don’t know the person? (It appears you acknowledged this in the same post.)

You are saying what it would require for one to be able to discern between a person and a machine. A characteristic of one person versus another person are just those behaviors that are repetitive and predictable.

I am confused: the point in time when a person becomes completely predictable to you is when he/she no longer has personality, as I have defined it. My definition of personality does not exclude the ability to be repetitive or predictable at all, but it does exclude being repetitive and predictable all the time.

An entity that always surprised you and was completely unpredictable would not resemble a real person at all.

Oftentimes, it is helpful to be able to note the two extremes and then understand the spectrum between them. Have I done that clearly enough by now?

As far as the Turing test goes, you have to take into account which human being your AI device is "attempting" to fool into believing he is communicating with another person. For instance, AI was advanced enough as early as 1982 to trick David Hasselhoff into having a personal relationship with his car K.I.T.T. I understand this required almost two dozen lines of code.

I question the claims that A) David Hasselhoff was tricked—versus acting his part, and B) two dozen lines of code in any known language, without use of complex libraries, could simulate an AI like K.I.T.T. (To continue this tangent, I would like to see sources for these claims.) I am a software developer and have familiarity of languages ranging from assembly to C/C++ to C# to LISP to Haskell, and even some HDL, including Verilog and VHDL. None of these allows what you seem to think they allow.

[cousinbasil]

As far as the Turing test goes, you have to take into account which human being your AI device is "attempting" to fool into believing he is communicating with another person. For instance, AI was advanced enough as early as 1982 to trick David Hasselhoff into having a personal relationship with his car K.I.T.T. I understand this required almost two dozen lines of code.

I question the claims that A) David Hasselhoff was tricked—versus acting his part, and B) two dozen lines of code in any known language, without use of complex libraries, could simulate an AI like K.I.T.T. (To continue this tangent, I would like to see sources for these claims.) I am a software developer and have familiarity of languages ranging from assembly to C/C++ to C# to LISP to Haskell, and even some HDL, including Verilog and VHDL. None of these allows what you seem to think they allow.

Uh... I was joking...?

[Luke Breuer]

Uh... I was joking...?

I'm sorry, sometimes I get locked into serious-mode...

[cousinbasil]

Uh... I was joking...?

I'm sorry, sometimes I get locked into serious-mode...

The point being, of course, that the Hoff is a pinhead and probably thinks a thermos is intelligent because it knows to keep hot things hot and cold things cold... In case you were still locked!!

[Luke Breuer]

Yes, yes, I am aware at poking fun at Hasselhoff. I find that much less interesting than the actual topic at hand. I have trained myself to gain nothing out of such mockery; perhaps I haven't yet to find the key where I am not building myself up by tearing someone else down. I will add a small personal note: many people have mocked me in that way and I made a solemn oath in my late teens that I would never intentionally treat others how I was treated. But this is a topic for another thread, so I stop here.

[overmyhead]

A system has personality if it has infinite unique variables associated with it. Perfect knowledge of such an entity is impossible through induction. Emotions (and perception in general?) arise from interference patterns between expected systems and actual systems. Is that the gist of it?

Your surprising conclusions are an artifact of the strange way you define "personality". It's strange to me that you use dry, removed academic language but choose to redefine such an emotionally laden word, rather than go with something like "autonomous" or "spontaneous".

I think your definitions are too specific. It seems to me that you are sticking with a formalism which you know should produce interference patterns. How about: something has personality if it is not perfectly knowable. Defining knowledge appropriately (e.g., knowledge is the absence of interference/ambiguity/confusion) allows you to start where I think you are really wanting to start: that interference, or something like it, is the basis of "_______" (I don't want to put words in your mouth).

A question for you: What is the difference between the interference pattern and personality itself?

[Luke Breuer]

A system has personality if it has infinite unique variables associated with it. Perfect knowledge of such an entity is impossible through induction. Emotions (and perception in general?) arise from interference patterns between expected systems and actual systems. Is that the gist of it?

I think you’re correct, including the idea that perception might be due to interference patterns. There’s also the idea of quantum decoherence, where amplitudes add instead of interfering; I haven’t yet worked out, philosophically, when one would happen versus the other.

Your surprising conclusions are an artifact of the strange way you define "personality". It's strange to me that you use dry, removed academic language but choose to redefine such an emotionally laden word, rather than go with something like "autonomous" or "spontaneous".

The word “autonomous” gives rise to “prime mover” discussions and ultimately, since we obviously don’t completely determine ourselves, the question of whether we retain any determining power. I think my definition is neutral with respect to this. The word “spontaneous” is ambiguous; in some senses it is allegedly “uncaused”, which is quite the dangerous claim, while in other senses, it is due to “randomness”; note that randomness is another way of saying, “I cannot detect any pattern”, not that “there is no pattern”. I think my definition is neutral with respect to this as well.

I think your definitions are too specific.

If I am right in that my definition is neutral with respect to “autonomy” and “spontaneity”, perhaps it is not too specific after all!

It seems to me that you are sticking with a formalism which you know should produce interference patterns.

The idea of interference patterns actually came after; it had no causative part as far as I know. The original formulation (which I posted on different forums) was actually to argue for reality having personality. :-)

How about: something has personality if it is not perfectly knowable.

What is not perfectly knowable, except for something that has infinite description? I suppose an alternative is that the thing has finite description, but you hit a wall, perhaps like you would hit an asymptote, whereby you can fully understand that last finite part. One way to differentiate between these alternatives is to A) well-order one’s attempts to know a thing, B) compare successive elements to see if one’s attempts are giving one less and less new information. Science is able to apply A quite well; B is a bit more of a judgment call. However, even if we are approaching an asymptote, I think my “interesting places” still remain.

I don’t like stopping at simple statements like “not perfectly knowable”, if there are options for more precision that, while they might be wrong, they also might lead to greater knowledge.

Defining knowledge appropriately (e.g., knowledge is the absence of interference/ambiguity/confusion) allows you to start where I think you are really wanting to start: that interference, or something like it, is the basis of "_______" (I don't want to put words in your mouth).

I’m simply not convinced that one can obtain “absolute knowledge” outside of a formal system where the “knowledge” takes the form of theorems, which can be proven from axioms. Do folks on these forums prefer using kinds of logic that cannot themselves be formalized?

A question for you: What is the difference between the interference pattern and personality itself?

Interference patterns can have voids (the waves cancel each other out) as well as changed amplitudes/phases. So, they can both lose information as well as distort information. This, I think, makes them very different from their constituent parts. That being said, there is nothing that says that interference patterns cannot have personality themselves. The requirement would be for the resultant interference pattern to still have infinite description.

[cousinbasil]

Yes, yes, I am aware at poking fun at Hasselhoff. I find that much less interesting than the actual topic at hand. I have trained myself to gain nothing out of such mockery; perhaps I haven't yet to find the key where I am not building myself up by tearing someone else down. I will add a small personal note: many people have mocked me in that way and I made a solemn oath in my late teens that I would never intentionally treat others how I was treated. But this is a topic for another thread, so I stop here.

It sounds like an admirable oath, and I hope you are sticking with it not only in refraining from mistreating others in the way you were mistreated, but in any way. In this case, to me at any rate, your oath seems a little too solemn. The Hoff earns his living by avidly seeking public reaction - he certainly is not within my sphere of recognition as the result of anything I have consciously done - so I have zero qualms about reacting to him or anyone else who has sought fame, especially those who chase the reality TV spotlight. Still, I can see a certain amount of humor might be lost on you, which is fine, we all have to put up with extraneous noise sometimes.

I take it that you are saying when a set has infinitely many recursively enumerable members, it is the same as saying countably infinite, like the set of natural numbers. I am a little unclear about this. Recursively enumerable has to imply countable, but I am not sure the reverse is true.

I’m simply not convinced that one can obtain “absolute knowledge” outside of a formal system where the “knowledge” takes the form of theorems, which can be proved from axioms. Do folks on these forums prefer using kinds of logic that cannot themselves be formalized?

I can only answer for myself. Formalizing logic is indispensable if one is concerned with proving the truth of an axiom or statement. This is using the definition that any statement in a formal system that can be proved true using only axioms of that system is itself a theorem. (Formal implying consistent).

It may very well be, however, that "absolute knowledge" cannot be thus shackled. If you require a formal, consistent system, you cannot also then require provability without sacrificing an unknown number of true statements - you seem to be well acquainted with Gödel and this was his insight. I am aware that technically what he demonstrated was that what such a system must contain is formally undecidable statements. But obviously one must conclude that some of these unprovable statements must be true, or else they would all be false and therefore would not be undecided.

You seem to be trying to define what constitutes a personality and what does not.

How about: something has personality if it is not perfectly knowable

Very few things are perfectly knowable, most of which cannot be said to have or be personalities.

Maybe you meant if it is perfectly knowable, then it cannot be a personality. Likewise one could observe, as I have earlier in this thread, if it is perfectly unknowable, it cannot be a personality. Unknowable seems to imply unrecognizable. We recognize personality traits all the time.

Personality might be much deeper than definitions or axioms or systems, and what we call someone's personality might be only that facet of it which we are recognizing at any given time. I think the clearest sense of one's own personality one ever gets is at those times when one feels misunderstood by literally everybody. Then one realizes that one is more than - other than - the sum of those facets.

Of course on this board, the reigning philosophy is that the ego is a delusion. But I think something of Das Ich survives when the facets shatter.

[Luke Breuer]

It sounds like an admirable oath, and I hope you are sticking with it not only in refraining from mistreating others in the way you were mistreated, but in any way.

Ahhh, but I cannot know every single action that could be considered “mistreating”. That requires A) proper teaching, and B) a tight feedback loop for when one makes mistakes. I don’t think A can ever be sufficient, so the honing of B is absolutely critical. These days, as long as someone is sufficiently precise, I can change my behavior after a single request.

I take it that you are saying when a set has infinitely many recursively enumerable members, it is the same as saying countably infinite, like the set of natural numbers. I am a little unclear about this. Recursively enumerable has to imply countable, but I am not sure the reverse is true.

The trick with countably infinite is that this restriction may relegate all of the elements to an ordered set which can be generated by a finite-length computer program. Perhaps the easiest way to think about this is whether you can take the infinity and use a super-fancy file compression tool to reduce it to a finite amount of data + a computer program. If you can do that, then your infinity is not the right kind for my definition. Looking at it another way, one can easily write a computer program that keeps on outputting data as long as you run it. Assuming it uses no entropy, no matter how much data it outputs, that data can be described finitely—by the computer program itself.

It may very well be, however, that "absolute knowledge" cannot be thus shackled. If you require a formal, consistent system, you cannot also then require provability without sacrificing an unknown number of true statements - you seem to be well acquainted with Gödel and this was his insight. I am aware that technically what he demonstrated was that what such a system must contain is formally undecidable statements. But obviously one must conclude that some of these unprovable statements must be true, or else they would all be false and therefore would not be undecided.

How can we tell the difference between hogwash and absolute knowledge, if no formal, consistent system is used for evaluation? “One just does.” sounds cultish and Gnostic, so I hope that is not your answer.

Very few things are perfectly knowable, most of which cannot be said to have or be personalities.

Maybe you meant if it is perfectly knowable, then it cannot be a personality.

What would you say is perfectly knowable, and yet has a personality? I am not sure any such entities exist—from any perspective other than God’s.

Likewise one could observe, as I have earlier in this thread, if it is perfectly unknowable, it cannot be a personality. Unknowable seems to imply unrecognizable.

I completely disagree, even if you say “unknowable in principle”. For one, how would one even know about something if it is perfectly unknowable? I’m not comfortable saying that humans must be the deciders of what has personality and what does not. If some entity seems unpredictable despite our best efforts, I think it is best to say, “I don’t know if he/she/it has a personality.”

Personality might be much deeper than definitions or axioms or systems, and what we call someone's personality might be only that facet of it which we are recognizing at any given time. I think the clearest sense of one's own personality one ever gets is at those times when one feels misunderstood by literally everybody. Then one realizes that one is more than - other than - the sum of those facets.

I never said that the sum of the recognized facets equals the infinite description. :-p This reminds me of the white stone in Revelation, which has a name that only God and the recipient know. I have been in the place of feeling misunderstood by everyone (which earned me the facet “persecution complex”); I decided to understand why, and came out of that knowing more about how people differ and how they communicate than most of my peers. Perhaps one of the most glorious things in the world is to help reveal new true facets of someone—facets which had not yet been shown to any human—in a safe environment.

[cousinbasil]

I’m not comfortable saying that humans must be the deciders of what has personality and what does not. If some entity seems unpredictable despite our best efforts, I think it is best to say, “I don’t know if he/she/it has a personality.”

Your condition of predictability vs. unpredictability seems insufficient on the face of it as being the sole criterion of which behavior indicates an underlying personality and which doesn't. You say you are not comfortable that humans must be the deciders of what has personality and what doesn't, yet the title you give your thread is "A definition of Personality" and you subsequently propose one. Presumably you are human, so I see a contradiction there.

My point is you cannot give a definition of personality (or anything else!) and exclude humans from being able to apply it. But I do think I know where you are trying to go with this: if personality (or perhaps a better term would be personhood) is an attribute of a "system," and one is going to discuss it in a meaningful way, one would like to have a working definition of it. For instance, is it an attribute a system either has or does not have, or can there be degrees of it?

Perhaps one of the most glorious things in the world is to help reveal new true facets of someone—facets which had not yet been shown to any human—in a safe environment.

Agreed - this requires the ability to love.

How can we tell the difference between hogwash and absolute knowledge, if no formal, consistent system is used for evaluation? “One just does.” sounds cultish and Gnostic, so I hope that is not your answer.

The key word in your objection in this objection is "0ne." One what?

[Luke Breuer]

Your condition of predictability vs. unpredictability seems insufficient on the face of it as being the sole criterion of which behavior indicates an underlying personality and which doesn't. You say you are not comfortable that humans must be the deciders of what has personality and what doesn't, yet the title you give your thread is "A definition of Personality" and you subsequently propose one. Presumably you are human, so I see a contradiction there.

I’m sorry; I meant to say that it is dangerous to say that because you cannot detect patterns, that there is none.

The key word in your objection in this objection is "0ne." One what?

“One”, as in “One goes to the market to buy food.”

[cousinbasil]

I’m sorry; I meant to say that it is dangerous to say that because you cannot detect patterns, that there is none.

It certainly isn't true that because you (one) cannot detect patterns that none exist. And in practice, it might very well be dangerous.

[Luke Breuer]

I think another inspiration for this definition of personality comes from subconscious (and now conscious) contemplation of Francis Schaeffer's discussion of a personal beginning of the universe in He is There and He is not Silent.

[paco]

fuck you

[cousinbasil]

fuck you

Happy New Year paco