Posts
Comments
I miss something about evolutionary game theory, where some of the discrepancies can be rationalized.
I wrote this tour from game theory to cultural evolution:
https://www.lesswrong.com/posts/xajeTjMtkGGEAwfbw/the-evolution-towards-the-blank-slate
We are surprisingly high in forebrain neuron count:
https://en.m.wikipedia.org/wiki/List_of_animals_by_number_of_neurons
Agree on this criticism for the difference between humans and pigs, but there too many orders of magnitude of difference between shrimp and human to consider detailed measures of computing power very necesary.
Quantifying empathy is intrinsically hard, because everything begins by postulating (not observing) consciousness in a group of beings, and that is only well grounded for humans. So, at the end, even if you are totally successful in developing a theory of human sentience, for other beings you are extrapolating. Anything beyond solipsism is a leap of faith (unlike you find St. Anselm ontological proof credible).
Illusionism is not a competitor, because consciousness is obviously an illusion. That is immediate since Descartes. That is why you cannot distinguish between "the true reality" and "matrix": both produce a legitimate stream of illusory experience ("you").
Epiphenomenalism is physicalist in the sense that it respects the autonomy and closeness of the physical world. Given that we are not p-zombis (because there is an "illusory" but immediate difference between real humans and p-zombies), that difference is precisely what we call “consciousness”.
Descartes+Laplace=Chalmers.
In fact, there is only one scape: consciousness could play an active role in the fundamental Laws of Physics. That would break the Descartes/Laplace orthogonality, making philosophy interesting again.
This is the kind of criticism I kindly welcome. I used the cockroach data (forebrain) here as a Proxy:
Thank you very much for the reference, because I am searching for co-authors for further develpments on SV-PAYW.
Also posted in EA Forum: https://forum.effectivealtruism.org/posts/uW77FSphM6yiMZTGg/why-not-parliamentarianism-book-by-tiago-ribeiro-dos-santos
That is the whole point of ethical systems, isn't it? To derive all (etical) values from a few postulates. Of course, most of valuations are not ethical (they are preferences or tastes), but this is an excellent agument for rational (systematic) Ethics.
Well, “one feel you can have done otherwise” is the part of the qualia of free will my definition do not legitimize.
When you chose among several options, the options are real (other person could have done otherwise) but once it is “you” who choses, mechanism imply “all degrees of freedom have been used”.
I say: "you are free when you do as you want, not matter how determined are your desires". This is how I define freedom in "Freedom under naturalistic dualism"(and I think that this position is original, so if this is not the case, I would be glad of being corrected).
You cannot know more than Laplace's demon, and the demon cannot assess consciousness. It is analyzed in detail in "Freedom under Naturalistic dualism".
Still is something that a conscient being superimpose over reality. It is not “there”. This as true for our moral as our mathematical constructions.
https://www.amazon.com/Morality-Mathematics-Justin-Clarke-Doane/dp/0198823665
This is also my intuition: the intensity of experience depends on the integrated information flow or the system and the nature of the experience depends on the software details.
Then iPhones have far more limited maximum intensity experience than ants, and ants maximum experience intensity is only a fraction of that of a mouse.
I mostly agree in the fact that while conscience intensity is the ontological basis of moral weights, there are other relevant layers. On the hand conscience looks to be some function of integrated information and computation in a network.
IIT for example suggests some entropic combinatorial measure, that very likely would explode.
In any case we are trapped in our own existence, so inter subjective comparison is both necessary and mostly depending on intuition.
What about an IPhone? It looks similar to a ant in terms of complexity; Less annoying too…
Because in the limit your intuition is that the experience of an electron is inexistent. The smaller the brain, the closer to inanimate matter.
Dear Jameson, as you say the theme is extremely important, but I miss more about Storable Votes: one period Arrovian results deeply change in dynamic voting scenarios. I have recently written two articles about this: one has been published in Journal of Economic Interaction and Coordination, the other is still a pre-print:
https://www.lesswrong.com/posts/5wqFoHBBgpdHeCLS6/storable-votes-with-a-pay-as-you-win-mechanism-a
I also suggest you to read the Casella and Mace review about “vote trading” (there a is Journal version, here you have the pre-print):
What I claim, is that with enough agenda setting manipulation you can nullify the properties of any voting system.
In my opinion, SV-PAYW is the best "voting system" available, but the mechanism has been analyzed under explicit hypoteses on the randomness of issues to be voted. The stream of political issues (represented by the valuation of the participants of an electoral victory) is supposed to be stochastic i.i.d.
If "deporting rationalists" is possible, and rationalists are not more than half of people, I don't see what security can they receive under any electoral system. If you can vote "disefranchise group X", then any minority group can be removed from the political system.
I was about to say that I explicitely deal with that issue on "the ideal political workflow", but there is nothing to deal with.
Well, if rationalists are a minority, with no external limits on the agenda, they can be deported anyway.
I only have considered a case with external agenda setting (issues with variable relevance exogenously arrive), as is typical to turn voting into a mathematical problem.
The second paper is about the context of voting systems. What I argue there is that the structure of the voting space is more important than the voting system.
What shall people vote? They shall vote among feasible states of the world.
This is the QV and I find it wrong. With money, at the end pivotal votes are only those of the largest money owners. Who can give money for collective choices? Those that can recover it, because their wealth is so large than individual consequences of collective decision is individually profitable. Keeping the political system separated from general purpose currency is critical (the Casella and Mace review agrees). The system as described in the paper is totally parallel to currency, while it works like it in the sense that you only pay votes when you get your alternative.
In SV PAYW the fixed number of votes is redistributed among voters after any election (votes casted in the winning alternative are those “payed”) in a one-vote one-man way. Even with no new votes for new players (and the votes from the dead are available for redistribution), new voters receive votes from the winners of each election.
A final remark: SV PAYW is more spectacular, but “the ideal political workflow” is more important.
People spend more votes on what they value more. In the original Casella system, every vote you cast, you lose it; my contribution was that you are only charged the votes casted in the winning alternative.
Absolutely incredible nobody suggested this before.
I still recommend you to read (and comment) the second pre- print.
Thank you to everybody for commenting!!
This is the whole point of the mechanism. To allocate victories to those who value them more. In the model there is a stochastic flow of issues with stochastic importance for both players.
The idea is that this system allocates victories to those who value them more.
How durable is money? In this version there is a fixed amount of votes circulating among voters, and votes can be stored indefinitely.
Of course, if this model were successful, versions with “storage costs” could be considered. Let 1000 flowers blossom!
I am no longer a gifted analyst, but I seriously doubt this. Differential games are well known, with their marvelous fixed point theorems and all the stuff.
In fact, if votes are not divisible you have to lot them, , as you can see in the (clumsy but tractable) discrete version I have analyzed in the paper.
SV PAYW is mainly designed for people to signal both intensity and direction of preferences. My opinion is that it is close to optimal to create conditions for truth telling of preferences.
In the situation considered in the paper there are only two players, so they know how many votes have themselves and the other player and the previous sequence of votes. The “incomplete information” situation means that the true value of wining in a given round is public.
While the voting system is very general, the situation considered is very simple, so recursive Nash equilibrium can be computed and simulated.
As commented in the second paper, unfortunately the big question is how to vote, but to create a meaning vote space… the question “what to vote” is in my view the most important. See “the ideal political workflow”.
"So it's compatibilist free will, and having a quale of libertarian free will doesn't change that"
Well, in fact the whole point was to legitimize the quale of libertarian free will (as a Hamiltonian, not as a unicorn), so I think at this point all our differences are reconciled :-)
"You have been assuming determinism, and if determinism is true you are not free to move your finger or not as you choose"
I freely chose to move my finger, but my (free) choice is pre-ordained. I can do as a I please (in the set of my "degrees of physical freedom"), so I am free within those bounds; on the other hand, what I want is pre-ordained.
"You have argued that some things that seem real are real because they seem real. The problem is that because the (degrees of) freedom don't really exist"
There are things I can do (because they are under my conscient control, like move my finger) and others that I cannot do (like altering Earth orbit). I am "free" to move or nor a finger, and I am not "free" to alter Earth orbit. Is there anything irrational in this? Anything illusory?
Of course, the way I use my freedom is determined. "Brain" has degrees of freedom, they are real (Laplace demon can compute them, by considering all futures conditional on possible Brain configurations), but, at the end, "Brain" is also a physical system, and can be predicted too.
The set of all possible futures conditional on brain configurations is the "degrees of freedom" of Brain, while, when you predict what "Brain" does with those degrees of freedom, you predict Brain use of its freedom.
How important is that dinstiction? From the perspective of Brain, hugue, for Laplace's demon irrelevant.
Both are rigth.
Now two smaller points:
"It's obvious to you that "all physical facts" doesn't include consciousness"
If Laplace demon needs conscience attribution to compute future positions and velocities, I rest my case. That could be possible, if "quantum collapse" is both ontologically real and triggered by a conscient observer. This is obviously over my payroll...
But what about ontologically? Surely that is the gold standard.
Regarding this, I have not better arguments than Bishop Berkeley, who turned the Cartesian epistemological subjectivism into ontological subjectivism. That is the real birth of anglo-continental divide, while perhaps Occam laid grounds to the tradition in the Middle Age.
That's the standard argument for compatibilist free will. Compatibilist free will only needs preferences and the ability to act on them. Adding a dualistic layer to physical reality doesn't give you any more compatibilist free will, and it doesn't get you libertarian free will, as you admit. So what is doing?
Let’s work over the case of a deterministic physicalist universe, to simplify. You can divide reality in two disjoint arbitrary sets covering the whole universe, called “Brain” and “Rest”. Laplace demon has all information about “Rest” while no information about “Brain”. Now, he can compute all possible futures of the universe for each possible configuration of “Brain”. This set of possible futures tells us the “degrees of freedom of Brain”, that is, how this subset of the reality can influence the Universe. This is a cumbersome but legitimate exercise, and a totally materialistic one.
Now, we go to the other side of reality: suppose that “Brain” is not an arbitrary part of the universe, but one that has an attached stream of unified conscientious experience (those systems exist: I am one of them, hopefully you too). “Brain” is volitive and has a feeling of free will and choice. What I say, is that the scope of possible futures among which a conscious being chose is an approximation to the “degrees of freedom of Brain” that I defined in the previous paragraph (with the help of a Laplace demon). What I claim from this is that the feeling and the scope of freedom are rigorous mental objects. That they illusory as a Hamiltonian, not as a unicorn!
So I guess my position is “compatibilist”, but my experience reading compatibilist texts is that they are not as clear as my exposition before.
Of course, everything depends on conscience, because it is the mental experience of free will what is under analysis here. What I am trying to do is to assess the rationality of that experience. The unpredictability of the disintegration of an atomic nucleus is not an instance of free will, and in fact, I have developed all my argumentation for a purely deterministic universe (but one where the future can be affected by the present, but no the other way around. How is that possible if the laws of Physics are time symmetric?… That is for me the real open question; the philosophical part [=relation between physical reality and epiphenomenal conscience] is straightforward).
We belive this, because Laplace demon has an explanation of all material facts (that is physicalism, isn't it?). What else can you "know", what else can you explain?
The most you can do is to trust is the "neural correlates of conscience" research agenda, but it depends on having a Rosetta stone (=credible accounts of subjective experience), and beyond other humans, we have nothing (while perhaps IA translation of some cetacea will be available, increasing a little bit the "interpretability circle").
But we will never know "What Is It Like to Be a Bat?"
“ What if there was an entity that could consider possible futures conditional on its actions and choose one of them - suiting its goals the best, - and yet this entity completely lacked conscious experience? What properties of free will would such entity lack? Or do you believe that such entity is impossible in our universe (or even in any universe?) thus leaving the possibility for the falsification of your theory?”
I don’t think I am making a “theory”, but more an “interpretation”.
My position is that freedom is a property related to conscious experience, so as long as there is no conscience, my definition of freedom cannot be applied. Conscious choice is the basis of freedom. In fact, this is specially obvious, because my definition of freedom opens the door to moral responsibility, that obviously cannot exist with no conscience! What am I missing here?
“Mental things are part of a map. Initially when you see a map you are more certain of its existence than the territory that this map describes. And yet the map itself can still be just a part of territory”
Well, this quite metaphorical for a hard materialist … :-)
Additionally, I cannot compare “map” and territory, because I (subject) can only deal with different maps (I cannot access reality, wherever it means, directly), at most different internal representations of something I believe is “out there”.
“Now you say that we can, if this something considers, chooses and has goals”
I am only sure of the fact that I "consider", "choose" and "have goals". For the rest, it is only a hypothesis (quite persuasive for very similar organisms that also speak, and the more different from me the harder is to use analogy to assess sentience).
“We have good understanding how to make one. We still lack in understanding how exactly it works. Interpretability research are ongoing and we are slowly learning more and more about it, filling the blanks in our understanding “
I agree with this. In this particular case of “biology”, we have perfect knowledge of biochemistry [=the generative model], but as humans we want also the intermediate layers (like cytology). Still, in my view it does not matter if you go top-down [classical biology] or bottom-up [AI interpretability], what you have is phenomenal knowledge.
The “neural correlates of conscience” people are working in axiomatic formulations (Information Integration Theory), because they can only trust human experience (where analogy and language are available). For the rest of beings we have no Rosetta Stone that allow sentience comparison. Even in the case of humans, everything depends on “trust” on others’ experience, because the only conscience I can measure is the mine one.
“It's not that you had the pleasure to experience the most perfect physicalist-reductionist explanation. So how can you know that you are not giving up too early? Epiphenomenalism doesn't try to solve the mystery so it won't be able to do it.”
Epiphenomenalism suggest that there is no mystery to solve. Of course, epiphenomenalists are as interested in “understanding” phenomenologically either AI and biological systems as anybody else. But after explaining everything, either by having the generative model of the systems or the intermediate layers of reduction, sentience can only be assessed by the the conscious entity itself.
Every conscious being in the universe is epistemologically alone, owning the knowledge of their sentience as a metaphysical absolute certainty that cannot be transferred to any other subject. Splendid loneliness...
Ape in the coat
“Are you using a different definition of "fundamental"? What I mean by it is an element that is not produced by any other elements”
Epistemologically, your subjective experience is more fundamental than physical reality. The entire world could be a simulation, and then Physics would be false; on the other hand, subjective experience and Mathematics is true in absolute and aprioristic grounds. Color red, an orgasm, and natural numbers and the Fermat theorem are true, even if you are trapped in Matrix, under the power of a “malign genius”.
Cartesian subjectivism and Newtonian mechanism were almost contemporaneous, and they created a massive Schism in philosophy and the Western mind. From the perspective of the mathematician and the philosopher, subjective experience is immediate and “fundamental”, while sensorial experience imply “faith”. You cannot prove anything about physical reality because you cannot prove there is a physical reality at all. But in Mathematics, you can prove many things, because they refer to mental objects, and your mind and its objects are (unproblematically) real. That is why mathematical theorems are more certain than physical laws. When you accept that, it implies that conscience is more epistemologically fundamental than physical reality. You are trapped in your self, and the “world” is not more than an act of faith or a useful hypothesis that coordinates your experience.
Of course, I have faith in sensory experience (unlike Tensor White, see comment below, that has more faith in God and less in brutal matter). But I am aware of the abyss that separates me from wherever Physical reality is, and the fact that that abyss can only be crossed by faith (unlike you accept St Anselm ontological proof of God existence, as Descartes did).
“Instead of trying to understand the mystery of how consciousness is produced by matter interactions it just postulates that it's a separate entity”
I begin my exposition describing what means “explanation” in Physicalism. Life is explained by pointing out that its apparent special characteristics (reproduction, autopoiesis, etc) are simply results of the laws of Physics. Reductionist biology takes a life form, makes inverse engeneering and show the “plans” of the being, and you can understand how it works from physical and chemical laws (that means finally only Physics).
But what about chat GTP? How helpful are the “plans” of the machine to assess conscience? What additional scientific knowledge do you need? You have the generative model, the most complete set of plans and physical relations you can dream of. Regarding chat GTP you are almost like the Laplace demon. Still, you don’t know how sentient is it.
For me that is “Naturalistic dualism”: the physicalist-reductionist explanation (no matter how perfect) is simply not enough for the assessment of sentience. Either there is some other type of scientific explanation beyond reductionism-physicalism, or simply conscience is beyond our scientific assessment (not understanding! The typical naturalistic dualist thinks that “there is nothing to explain” nor understand, probably because we accept what explanation is since Newtonian “hypothsys non fingo”). Sentience is real, and it is simply “there” for some physical systems, there is no way to assess which ones.
“What if there was an entity that could consider possible futures conditional on its actions and choose one of them - suiting its goals the best, - and yet this entity completely lacked conscious experience?”
If it “considers”, “chose” and “has goals”, it is conscious.
Tensor White
Thank you for your comment. I cannot answer you with the kind of detail that I did to “Ape in the coat”, because I am a materialist (as materialist as possible, in a Universe where conscience exists). I agree that it takes a lot of intellectual machinery and some leaps of faith, but reality is as it is: atoms and emptiness, as Democritus discovered more than 2500 years ago.
TAG
“Doesnt that amount to saying that free will isn't really real,but seems real?”
It is as “real” as the set of natural numbers or as real as “true love”. For me that is more real than a neutron star that nobody has ever seen. The stream of my conscience is more real than anything else and well defined mental objects there are very real.
“Libertarian free will allows the future to depend on decisions which are not themselves determined.”
Everything is physically determined. But when a conscious being has N options, and choses one, he chooses what he wants (=he is free), no matter how determined is what he wants.
The discussion on moral responsibility comes from that: when you accidentally kill your wife you can say you didn’t chose. On the other hand, if there is intent, the determination by material causes of your act does not change that the act was conscious and then your conscience [=you] is evil.
As Jan Blomqvist said, we are “ghosted machines”. The fact that the ghost comes from the machine doesn’t mean that the product of the machine cannot be an evil ghost.
Thank you very much for this marvelous discussion. At least, even if does not end as an academic paper, I am very happy with this exchange.
Sorry for the delay, but there are posting (one per hour, 3 per day) limitations for low karma participants even in their own posts.
I answer all comments together (with low karma you can only make one comment by hour):
"Consciousness doesn't need to be fundamentally distinct from non-consciousness. Rocks can't monitor their own states at all, but computers can, that doesn't mean that a fundamentally new property was added when you turn a rock into a computer. ". "Consciousness not being fundamental doesn't equal consciousness not existing."
Conscience is "fundamental" in the sense that the philosopher itself is the conscious subject. In fact, the entire physical reality could be unreal, and still the “brain on the vat” would be real and its experience too. The self is immediately real, and that is why Descartes is the father of modern philosophy. If I were a “brain on a vat”, still there would be an infinity number of primes, and pleasure and pain would be entirely real. In that sense, conscience is fundamental: in the sense that it is the "hard" reality for you (subject) no matter what causes it. On the other hand, consciousness is not "fundamental" in naturalistic dualism in the sense that it does not even play any role in physical reality: it is only epiphenomenal.
In our current physicalist-reductionist vision of Nature, the Laplace demon only deals with positions and speeds (or more exactly wave functions) of all particles in the universe. This is a complete and autonomous description of Nature, and being complete and not having "conscience" as an input, the demon cannot assess the "other side" of reality (the side where Descartes thinks and consequently is). Perhaps, if quantum wave collapse is related to conscience, then, conscience could play an active role in reality and this article would have to re-written after some additional courses on Physics. Otherwise epifenomenality is un-avoidable.
"Try removing the premise of dualism from your reasoning"
How dualistic is "Naturalistic dualism" when conscience is purely passive and epiphenomenal? In my opinion, naturalistic dualism is the most materialistic non-eliminativist worldview (and eliminativism is meaningless).
"Another direction in which you may want to look is distincting free will from counsciousness"
In fact, "free will" in my definition is some very specific consequence of conscience and time asymmetry. The fact that the physical system with an attached conscience flow can "affect" reality [=possible Worlds conditional on own body actions], means that it considers some possible futures and choses one of them. The possible futures given own choice are meaningful, but for an external observer that can predict also the evolution of the conscious being [=brain], the choice is as materially determined as the rest of the Universe. The existence of conscience is what divides the universe on two subsistems, and makes sensible to use "own possible actions on reality" as free variable to pursue (volitional) ends. So in my proposal, free will is not confused with conscience!
"But conscious states are strongly determined by brain states as far as we can check"
You can only "check" your own mental states, so that is not very far.
"Consciousness doesn't need to be fundamentally distinct from non-consciousness"
I cannot argue against eliminativism, because perhaps you are nos conscious. Still, I would not eat you because of cultural taboos and legal complications...but no longer for moral reasons! :-)
As commented in the article, this philosopher is conscious for sure, but the regarding the gentle reader, he only can hope. Not even the Laplace demon would know, and my phenomenic knowledge is vastly inferior.
The problem with conscienciouness, is that full scientific knowledge of a phenomenon, tells nothing about it. It has been allways obvious, but now, with neural networks it is even more evident. You have the generative model, that is, the perfect scientic knowledge on a system. Still you know nothing about sentience.
I agree with Chalmers, but I dislike his presentation of dualistic naturalism, because he goes into long mental experiments. This is far more immediate: Laplace demon cannot assess sentience. And Laplace's deamon is the "omniscient materialist"... In this limited (but critical) sense, conscience is not material.
You dont need to postulate conterfactual zombies.
Dear all,
I will keep any remaining discussion on the EA Forum. It is the version of the article that has been commented in Marginal Revolution (point 6, second link):
https://marginalrevolution.com/marginalrevolution/2023/04/thursday-assorted-links-400.html
Nuclear winter:
https://climate.envsci.rutgers.edu/pdf/WiresClimateChangeNW.pdf
Electromagnetic pulse would destroy most electronic devices:
https://doh.wa.gov/sites/default/files/legacy/Documents/Pubs/320-090_elecpuls_fs.pdf
"Commercial computer equipment is particularly vulnerable to EMP effects. Computers used in data processing systems, communications systems, displays, industrial control applications, including road and rail signaling, and those embedded in military equipment, such as signal processors, electronic flight controls and digital engine control systems, are all potentially vulnerable to the EMP effect. Other electronic devices and electrical equipment may also be destroyed by the EMP effect. Telecommunications equipment can be highly vulnerable and receivers of all varieties are particularly sensitive to EMP. Therefore radar and electronic warfare equipment, satellite, microwave, UHF, VHF, HF and low band communications equipment and television equipment are all potentially vulnerable to the EMP effect. Cars with electronic ignition systems/ and ignition chips are also vulnerable."
not a single open society would survive a nuclear war; nuclear war inevitably leads to more nuclear war; nuclear war will cause widespread societal collapse
Do you expect a paper? I have this one:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC300808/
:-) After total economic disruption, no electricity nor electronics, and a few billion deaths... its like the parachute randomized trial. Too obvious to be argued.
Now let´s compare with peer reviewed literature on AGI:
https://marginalrevolution.com/marginalrevolution/2023/04/from-the-comments-on-ai-safety.html
"The only peer-reviewed paper making the case for AI risk that I know of is: https://onlinelibrary.wiley.com/doi/10.1002/aaai.12064. Though note that my paper (the second you linked) is currently under review at a top ML conference."
But what is the probability that AGI wipes us? Why would AGI be more aggresive than humans? Specially if we carefully nurture her to be our Queen!
Why do you think AGI would necessarily be worse than us? I think we really don't know.
We would not be destroyed in the first large nuclear war (while effects of radioactivity in the food chains in my view are under researched). But not a single open society would survive.
A world of malthusian masses, and a military aristocracy desperately trying to keep as much firepower as possible is the natural post nuclear war outcome. Then, history happens again, and in 2000 years luckily we are back into deciding if we allow AGI to be developed. What is the point?
The baseline is that our governance systems are completely unaccurate for nuclear weapons. Even in this fortunate age of hegemonic republics.
We need to solve the human alignement problem. Do you have any better suggestion than AGI?
Well, after a complete NATO Russia exchange, direct deaths would be in dozens millions in the first week, and the electromagnetic pulse would left the power production systems and the majority of electronics destroyed.
On top of that, you have nuclear winter, that put the deaths in the billions (see link in the text).
And then what social system is left? A second wave of wars would be inevitable, and inevitably nuclear.
Well this answer to the "individual well being"; the equal weigths are worth discussing. Do you have any suggestion for different weights?
For ontological reasons! Value is something that consciece attributes to a dead cold indifferent universe. Value is subjetive as mass is inertial. Too fundamental to be discussed, I would say.
I would say in philosophy, but not very much in political philosophy. The "Social welfare" definition is more Ethical than political, the rest is purely "game theoretical". But of course, at the bottom everything has metaphysical hypotheses. The less and more general, the better!