Love as a universal concept

Love is something every human being knows as non verbal knowledge, but when we define systems and machines with some form of intelligence (strong AI) we also need to define this in a stringent terminology that can be represented and implemented in algorithms and behaviors.

First, the concept of love is even in humans somewhat ambiguous, and can be broken down into:
0. mutual love, which can be considered, love by contract.
1. agape or unconditional love.

Are both of these essential for a machine? In my view yes, especially if we speak about autonomous robust entities which could be dropped down into any type of scenario and solve problems within that scenario, by creating an internal list of problems that need to be solved, then prioritize these and solve them in some order of significance, ability and causality.

Long time ago, in March 2000, during my PhD program after some pondering over a specific problem I scribbled down an introspective approach to an ethical AI algorithm based upon love, this is in the speculative part of my thesis (ch 7), and also as a brief slashdot comment here.

Then, at a conference about nano technology I in Palo Alto in April 2004, organized by Foresight Institute I attended a workshop about safe AI, led by  Steve Omohundro where there were around 25 strong AI researchers present, we discussed the problem of creating safe AI. I proposed love as a fundamental concept and reached a consensus among the audience that this is it.

Now, the problem is that “love” is considered an ill defined concept as it also need to be formalized in an axiomatic or mathematical form, which can be understood by the machine, and so far I haven’t seen any strict definition of the concept of love.

Let’s start with the unconditional love, which is usually less understood by humans, but I claim that this is the easy part, as it can be defined in a strict manner, where mutual love needs dynamic programming.

My simple proposed definition of unconditional love, for any system:

Strive for holistic consistency.

Here as English is somewhat ambiguous, “holistic” simply means: look upon the whole context, that is, don’t deliberately reject any theorem or information.
Regarding consistency, that has a strict meaning in technical and mathematical terms. A system is consistent if it doesn’t contain contradictions, as in a system which contains contradictions, anything can be proven as truth. A consistent system in mathematical terms can then simply be considered a “true” system.

In engineering (technical/social/economics/politics/software etc) it simply means a system with conflict free solutions, that is, a solution where one part of the system is not trying to beat another part of the system (not in a competitive way, that is different). Therefore, the strive for holistic consistency could be seen as a goal generator, that is allowing the system to identify the problems in the system, without explicit programming.

Then the agents (term for autonomous systems with a specific agenda in technical terminology) need of course to interact with each other and with other beings. In social contexts there is a well established rule denoted “The Golden Rule“treat others as they want/need to be treated“. Observe, this is not the standard definition, which is treat others as you would want to be treated. However, the latter is ill defined in a sense as not every being like to be treated the same. My own approach to this uses dynamic programming:

while true do:

    1. Treat others in the way you would like to be treated as first approach.
    2. If they respond by being rude, then respond by being somewhat less rude.
    3. else if they respond by being good, then respond by being somewhat less good (i.e. do not compete or exaggerate).

The process is repeated forever and (usually) reaches a dynamic mutual balance, where you over time may understand contextual dependencies. Now it is very useful to discuss issues and in a discussion about strong AI on facebook recently I got a suggestion from AI researcher Mark Waser that this is an extended version of what in game theoretical contexts is denoted Optimistic Tit for tat and, yes, I agree upon this.
The old Tit for tat has equal retaliation and does not encourage collaboration. What is an extension is that this behavior strives for quicker balance as it has a weak retaliation (I consider that “revenge” creates unstable solutions). This type of game theoretical models can solve tricky scenarios like the prisoner’s dilemma, and I consider that this converges towards a Nash equilibrium (John Forbes Nash got the Nobel Prize 1994 in Economical Sciences).

It should be noted (thanks Mikael Djurfeldt) that if we would only strive for holistic consistency, one solution is an empty world, where there can be no contradictions.

It’s only when these two definitions of love taken together it creates the condition for a being to strive for non empty worlds, as mutual love, i.e. the strive to treat others as they want to be treated, creates a motivation to strive for having someone to treat.

Thus: Love as a driving fundamental force could be summarized as (thanks Eray Özkural for helping me realize this principle):

0. strive for having someone/something to treat/nurture.
1. holistic consistency.

For my own I like metaphors, and for a computer scientist , a natural approach may be to use software licenses as metaphors, then I propose e.g (I’m aware that not all people may agree upon this metaphor):

Proprietary (closed source) Evil, as it creates a non productive asymmetry in the system
GPL/CC-SA/copyleft Mutual Love, that is love by contract
BSD/Public domain Agape, that is unconditional love

This could also imply that if there were no evil proprietary software, there would not be any need for the copyleft, love by contract version, it could be enough with the one based upon unconditional love. However, as the unconditional love is not a sufficient condition [remark added Dec 25th], the mutual love builtin in GPL/copyleft creates a condition for all beings to strive for the common good.

Regarding love by contract, i.e. which applies also to generic products and product developing we have generalized the concept of free software to free computer (it’s the AI frame work for this we are developing).

It should also be noted, that this simple formal definition of love above, of course leaves out many meanings of love which includes passion, strive for experiencing beauty as music and art etc, this definition only attempts to define the sufficient conditions for any being to be collaborative.

Now, as we are all in some sense in a type of prison, that is our world has borders, which can be illustrated by this picture:

Then, an interesting issue is how this “holistic consistency” relates to Gödel’s incompleteness theorem (the axioms within a system can not prove the system to be consistent), I’ll later (as I at the moment need to assist with other stuff) explore this in more detail. Many people have this view (thanks David Jansson) about their freedom (and many actors in the society tend to implement this in different ways…):

Freedom

However, I claim, that most intelligent agents would prefer to be see their freedom like this:

I claim that this is possible in all systems, but the system then needs to include a self supervision (the eye) to guarantee it’s own consistency.

I’ll ponder over this further soon, but that’s all for today.
Best holiday wishes/Roland

PS. extremely happy for any type of comments and feedback on this.

15 thoughts on “Love as a universal concept

  1. You begin with the correct observation that love is ambiguous but then resolves than by a simple division. I don’t think you can do a meaningful reduction of the concept of “love”. Certainly there are aspects to it and each aspect can be viewed from a specific context however love tend to be more than the sum of its parts. Just as the solution for the prisoners dilemma changes as time progress and even the expectation of future time changes, the expectation on love will change over time based on the experienced love.

    Further more I believe that it is possible to commit reckless or even evil acts based on ‘love’. It is not good enough for a “safe AI”. The safety you want from a thinking machine is an easily accessible turn-off switch or another thinking machine set to watch the first one.

    • I agree to some extent. First the separation in unconditional and conditional love I consider a necessity as it is otherwise not easy to make a simple algorithm ;) but the great thing with this separation is that the consistency check can also be seen as both a supervisory model of its own system as well as a goal generator. Then the mutual love part is certainly much more complex in a more general perspective, although it’s not for sure that we want it to have all.

      Then the on/off switch you suggest is also dangerous, sure there are no completely safe systems, but this simple algorithm at least has the advantage of being much safer than any human, as humans also have emotional irrational responses that can be dangerous. Also, if the AI has an on/off switch, how can you trust the being turning it off? The AI was just going to save the humanity, but due to the on/off switch it couldn’t.

      The bigger issue is though: Do we really want or need this type av more personalized AI as some kind of robots?
      This one is such a simple model though that I think it may have many usages, like as an assistant for quick access to complex operations, but an interesting issue is, will this type of quite simple algorithm evolve a kind of consciousness? It is “self aware” by definition, merely as a part of a control loop to be able to plan its own actions.

      The problem with “it is possible to commit reckless or even evil acts based on ‘love’” sure, but now I do think that you relate to several emotional actions humans can do in the act of “love” even though it may even be considered a revenge action. This AI won’t ever think in terms of “revenge”. That is, a human “emotion” which is not really productive I would say.

      Also, as this is quite simple (OK, the bottom layers are not) it is of course simple to do experiments with. The tricky part though is the consistency check which is a somewhat ill defined problem. My idea is that it learns the local context and can detect local problems, then over time will learn more and more, until it can detect global or even more universal problems.

      For all practical purposes, I would say that when developing this, one do it incrementally, that is start with the problem generator, the consistency check, which then can trigger a message response, to tell someone about the problem. It’s probably not wise to develop it with full actuators and things like that from the beginning.

  2. ” The tricky part though is the consistency check which is a somewhat ill defined problem”

    You are referring to “holistic consistency” here? I think you need to elaborate more on that. As you probably noted, Monica Andersson has rigorously ironed out what the holistic approach means and how it differs from a reductionist approach. In contrast last time I chatted with Eray Özkural he despised anything non-reductionistic and seemed to be obsessing about the existence of a singularity. (in fact he took criticism very hard)

    The second attempt to reduce love by simple division is also still a puzzle to me:
    “0. strive for having someone/something to treat/nurture.
    1. holistic consistency.”

    Item 0 is an generalized observation on a behavior. It isn’t an attribute of ‘love’ itself. Item 1 isn’t clarified what it means. I disagree that contradiction is an issue for human like intelligence. Many humans happily accept contradictions in their life. Contradictions is something you want to avoid if you are doing rational thinking, which is something that emerge from human intelligence. Human intelligence in turn emerge from sub-intelligent processes.

    Instead of using the love metaphor above I would have looked at the different kind of emerging behaviors we observe in competitive evolution:
    http://en.wikipedia.org/wiki/Mutualism_(biology)

    It presents both more dimensions to your classification attempt and it is also more or less established science. Then at least you can reduce the vocabulary to a known set.

    • First, to not induce some conflict or competitive thought trains. My own brain works as a hypothesis “Borg” that is assimilates all possible ideas and concepts and see them just as hypotheses, and I don’t believe in what anyone else says in any other meaning than creating hypothesis which I myself can correlate with observables (like own significant observations or similar scientific observations from several independent sources).

      I haven’t read what Monica Andersson has defined, for me it was enough that she stated the question about holism versus reductionism (as I’ve always considered myself a holist), they are both essential, you can not rely on only one of them. Then you wouldn’t be able to solve anything, to come to any meaningful conclusion. If you are familiar to backtracking strategies in e.g. systems working with predicate logic I consider the holistic view is “breadth first” and the opposite not really to be reductionism, but inductive reasoning, that is you build upon what you already know. As a human metaphor, although I’m not much into categorizing people, I would claim that “holistic” view is (according to inflicted stereotypes) a woman’s view and “inductive reasoner’s” a man’s view, now not speaking about the reproductive apparatus which makes things more complicated… (as it becomes a contextual reversal in that sense).

      Regarding “holistic consistence” it is a well defined concept in logical reasoning, that is a system containing no conflicting solutions, and I also claim (I’ll come back to that in a later contribution) that our brains are actually capable of exactly this, by implementing both implication logic reasoning (thinking) and prediction logic (common sense). Neuroscience so far has focused on the pattern recognition part which I would claim is only solving a subset of the problem (forming of feature detectors, segmentation and forming of concepts).

      Regarding my love ’0′ : Yes, this is purely based upon observations upon behaviors. I’m not sure I understand your “isn’t an attribute of ‘love’ itself as then it becomes a circular definition based upon “love” which is exactly what I’m trying to define (in machine implementable form, and I claim that this is enough, a machine built upon these two definitions will become intelligent).
      Regarding my love ’1′: As I said, I considered this the easy part, but it is still not very well understood, but as I see it is as simple as strive for non contradictive solutions.

      Regarding: “Many humans happily accept contradictions in their life.” I don’t!
      And in an extended context accepting these contradictions leads to a crazy system. One thing for instance, which I consider important, people seem often to mix up the “concept/class” with the “object/instantiation”, which leads to a crazy behavior.

      Take for instance something like lies. Why do we have lies? Why are there people that lie? as lies creates internal conflicts as well as larger conflicts.

      I see the reason why lies exist merely as a side effect of a defense system that can be motivated as long as people see each other as enemies and they want to defend their position, however lies (which a large part of our society is built upon today) creates a non working system.

      There are “lies” and there are “white lies”. For my own I see the difference as “lies” induces conflicts and “white lies” do not.

      Observe when I speak about conflicts I speak about internal conflicts “brains” and external conflicts “world”.

      And, regarding your reference to “mutualism” that is exactly my love 0. But, if you only see things outgoing from your self, I consider that this induces conflicts at the system level which leads to an unstable or highly “unfair” system, with lots of conflicts.

      • ” If you are familiar to backtracking strategies in e.g. systems working with predicate logic I consider the holistic view is “breadth first” and the opposite not really to be reductionism, but inductive reasoning, that is you build upon what you already know.” => I see. This is far from Monicas contribution to AGI though. It might be interesting for you to digest her videos just because of the radical direction she takes compared to the description above: http://videos.syntience.com/ai-meetups/andiair.html

        “[..] that our brains are actually capable of exactly this, by implementing both implication logic reasoning (thinking) and prediction logic (common sense). ” => I think they are, but I think it is an emerging effect from underlying structures which doesn’t “think” at all, i.e. they are sub-intelligent and use neither inductive nor deductive reasoning. In fact I’m quite convinced the brain happily adopts ad-hoc ideas based on previous experience, significant events in that experience. The restriction for how the sub-intelligence networks are shaped is influenced by biological contracts from inner organs.

        I disagree logic has anything to do with common sense, which to me rather is an expression for a mean value based on previous experience. It is common sense that the light bulb turns on when I flip the switch because it has always (to my knowledge) done so when I’ve tried. It is not a resolution of predicate logic which gives me this idea but a combination of inbound signals forming an index pointing me to previous experiences and exformation removing insignificant data. As new actions affecting the world are defined by both new observation and previous experience is room for a strange attractor creating non-deterministic behavior in the thought process despite the simplicity in the lower level logic.

        I don’t see any evidence that a thought process reduces the world around it to find solution to ‘problems’. In fact I find no evidence of generic ‘problems’ at all. From a self-centered perspective I know that pain and sometimes pleasure forces me to act in a certain way, and when that way becomes hard I call it a ‘problem’. Hence I don’t think logic or reasoning is important at all for AGI. Logic is certainly important for any human living in our culture today, but hardly for a bear, a dolphin or well.. let say a gladiator fighting in Rome in 300 AC. Logic and reduction are tools aiding our intelligence by cutting down and transforming past data and enable us to digest it when it is more suitable for us (i.e. it fits our culture and past knowledge better).

        From this follows that I can’t agree on why people lie either. A lie has no conceptual meaning to the sub-intelligent networks which makes up the foundation on which they are built. As you pointed out there is no point in reasoning in lies because they usually end up in not-so-useful contradictions. The reason you lie is because past experience and biological cursors favor this behavior and it gives you pleasure or sooth pain. Can it be regarded as a “defense system”? Maybe. But it would be subset of all its common applications.

        “But, if you only see things outgoing from your self, I consider that this induces conflicts at the system level which leads to an unstable or highly “unfair” system, with lots of conflicts.” => I don’t get this last part. You claim I only see things outgoing from myself? Maybe, how so? There are conflicts in my life, and contradiction but not to a degree I would denote it as a ‘problem’. Maybe you can explain more what you mean here?

  3. This what you said here:
    “In fact I’m quite convinced the brain happily adopts ad-hoc ideas based on previous experience, significant events in that experience.”

    is exactly what I see as the problem. People are listening to what other people say, which has made their logical thinking blunt, due to exactly the problem that people listen to what other people say, and think what they say is the truth, instead of seeing what other people say as a basis for concepts and hypothesis generation, which leads to such a crazy world we have today.

    Also you said: “I disagree logic has anything to do with common sense”
    yes, I never said that. Logic is logic, and common sense is prediction based upon logic and experiences. I have written a new contribution about “free will” which will try to exemplify this.

  4. Regarding: “But, if you only see things outgoing from your self, I consider that this induces conflicts at the system level which leads to an unstable or highly “unfair” system, with lots of conflicts.”

    Well, this is obvious if you look upon the world of today. The capitalistic system, the most insane religion I can imagine, is based upon flawed thinking. People (and companies) strive to get it good for themselves, without seeing what type of conflicts this system introduces, which maintains wars, poverty, stupid jobs which has no actual relevance and blunts people’s ability to think.
    A system which is not based upon logic (consistency) is a psychotic system.

  5. Can you prove, using logic, that the “capitalistic system is broken”? I seriously doubt it. Anyway, there is no doubt that logic is a great tool for creating consistency in thought and to organize thoughts so that you maximize the probability for desired outcome. The desires even, stem from processes which are ‘logical’. If you could freeze the world and analyze every dynamics you could predict what desires you would perceive next. Unfortunately you can’t make time stand still and the equations you’ll have to solve will have expired when you arrive with your answer regardless how fast you calculate. This excellent video by professor Sapolsky I think brings light to these issues: http://www.youtube.com/watch?v=_njf8jwEGRo&feature=related

    Logic and mathematics aren’t that hard. If we could simply take our observations and put them in a function and arrive with the purpose of life, we would. We can’t. Because of logic.

    You are worried that because people are looking at the world from within their own perspective they make shortsighted decisions. I think by assuming that if people made logical decisions you are making your self guilty of exactly they same behavior. You assume that you know what parameters other people are acting by and you assume that there is someway to get all significant parameters at high enough accuracy. Life is psychotic and chaotic. Consistency isn’t life. That is why nature evolve a brains before it evolved a binary computer capable of solving simple differential equations.

    • You ask: ‘Can you prove, using logic, that the “capitalistic system is broken”?

      It depends on how you define “proof”. I claim, and most mathematicians and logicians would agree upon this:
      A system which induces conflicts which contradict the purpose of the system, contradicts the purpose of the system itself.

      Now, you can not prove a system to be complete mathematically, (Gödel’s incompleteness theorem), but I claim that this can be done by combining logics with math and probabilistic reasoning, which is, what I claim, what intelligence is basically doing. However, regarding the capitalistic system, I consider the contradiction obvious, so there it can be proven.

      Interesting that you mention “meaning of life” as that was exactly what I intended to describe today. However, it is a function that can not be defined by math, it can not be defied by logic, it can not be defined by philosophy, fiction nor religion. The “meaning of life” is just to define yourself, to try to prove your own existence (from a reductionistic view point, although I claim to not be a reductionist :)

      It’s only when you combine all these, then you can do that. I’ll not try to prove it, proofs were never my strong side, my math teacher denoted my proofs “persuasion attempts” ;)

      However, when you tell someone something, you can never tell the whole truth, as the whole truth can not be known. I consider the best thing one can do is to provide a clue, and make someone think, thus grow themselves.

      You say:
      “I think by assuming that if people made logical decisions you are making your self guilty of exactly they same behavior.”

      Yes of course, if I asked people to believe me, which I don’t, as that would contradict my own thesis: don’t believe what people say :) . I just want to make people start to think, to create themselves, to understand themselves.

  6. Spot on with this write-up, I actually assume this web site needs a lot much more consideration. I’ll probably be once a lot more to read far much more, thanks for that information.

  7. Aw, this is an incredibly good post. In thought I would like to put in spot writing like this moreover – spending time and actual effort to create a great article… but exactly what do I say… I procrastinate alot by way of no indicates seem to get something accomplished.

  8. An impressive share, I just given this onto a colleague who was simply doing small analysis with this. And the man in reality bought me breakfast because I discovered it for him.. smile. So permit me to reword that: Thnx with the treat! But yeah Thnkx for spending plenty of time to go more than this, I’m strongly concerning this and enjoy reading considerably a lot more about this topic. If at all possible, as you become expertise, do you mind updating your website with an increase of details? It actually is extremely valuable for me. Huge thumb up because of this post!

Leave a Reply