A Sliver of Reality
A thought provoking article by David H Wolpert A Sliver of Reality
The question of what we can know of that which lies beyond the limits of our imagination is partially about the biological function of intelligence, and partially about our greatest cognitive prostheses, particularly human language and mathematics. Its also about the possibility of a physical reality that far exceeds our own, or endless simulated realities running in the computers of advanced nonhuman lifeforms. And its about our technological progeny, those children who will one day cognitively eclipse us. From the perspective of my 10 queries, human exceptionalism becomes very shaky.
Comments (22)
Wolperts concept of the future of knowledge is that which lies beyond the limits of our imagination in a real world waiting for us to represent it. I believe instead
that when we conjure the concept of a future we are always referring to what lies AT the limits of our imagination. A physical reality can never far exceed our own , given that physical reality is the set of goal-oriented interactive performances of humans on our environment , coupled with the feedback and constraints our performances induce. We might instead talk about how our future constructions of the world may be quite different from current ones, while avoiding decoupling that world from our interaction with it. From this vantage perhaps we can see that the issue of human exceptionalism originates from a confused idea of the relation between ourselves and our devices. Our devices, like our world , can never be beyond us. Our machines belong to our embodiment, as any other appendage.
This is not a given, it is a concept of reality that you endorse. It misses the point. It is as if a dog or paramecium (his examples) claiming that physical reality is the set of goal-oriented interactive performances of dogs or paramecium on their environment. The fact that we can entertain and express such ideas does not mean that reality is limited to what we think and do. It may be only "a sliver of reality".
Quoting Joshs
Our devices may someday be beyond us. In some ways they already are. Our world is not of our making and not in our control. It is not ours in that it controls the shots and we have limited power to change that. The world does not answer to us.
The scientific world we live in is a responsive order. The world responds very precisely, but differently , to the different formulations we impose on it , and has no sense whatsoever outside of all formulations. This is why what constitutes evidence in one model may be invisible within another model. We dont unilaterally call the shots and neither does our world. The responsive order is one of reciprocal causation between conceptualization and world.
A dog or paramecium constructs a niche based on its purposes as a self-organizing living system, and is in turn responsive and answerable to the effects of its efforts on that niche. What lies outside of that organism-environment responsive order is irrelevant to the organism. It makes no sense to talk about what might affect the organism from the world outside of its ow. making (niche) because there is no world for an organism outside of its reciprocally interactive niche. The same is true of our sciences. There has never been , and there never can be, a world for us outside of the continually evolving response theoretical niches we construct in interaction with that aspect of world that is relevant to our goals and purposes.
Wolpeet gives the impression the world can break through from outside this reciprocally responsive interaction to affect us directly, but if it did it would be invisible and irrelevant to us.
Quoting Fooloso4
Saying our machine are smarter or dumber than us is like saying the spider web or birds nest is smarter or dumber than the spider or bird. Should not these extensions of the animal be considered a part of the living system? When an animal constructs a niche it isnt inventing a life-form, it is enacting and articulating its own life form. Machines, as parts of niches , belong intimately and inextricably to the living self-organizing systems that we are.
Provincialism dressed up in sophisticated clothing is still provincialism.
Quoting Joshs
What does he say to give you that impression? It is not a matter of the world breaking through but of our expanding what we know. That has limits, but they are our limits not the limits of reality.
Quoting Joshs
Poor analogy. Spider webs and bird nests are not capable of self-learning or self-improvement.
This is a powerful image. Is there any sense to you that our continually evolving response is in some way inevitable based perhaps on the intrinsic limitations and opportunities inherent in being? By the way, I'm trying to ask this question without trapping myself in notions of time or destiny.
They are not capable of anything beyond our models which produce them, and these technological
models are themselves less advanced than our most speculative thinking, which is what makes the computer an appendage, just as our theories are appendages.
A computer capable of self-learning is able to do more than the programs that produce them.In addition, they are capable of doing what we are not. In any case, the article is about the limits of human knowledge, not IA.
He says this explicitly, as do you , when you both make a distinction between our limits and the limits of reality. Reality breaks through as something completely outside of our space of reasons, rather than emerging from within our shifting space of reasons.
If you are calling the advances in knowledge over time the world breaking through then I see no problem, but if you mean "reciprocally responsive interaction" I don't see where he makes such a claim.
I don't recall him saying anything about the limits of reality.
Quoting Fooloso4
Maybe its about both. Wolpert seems to consider AI as the solution to the limitations of human knowledge. He concludes the paper with speculation about future members of any species that we consciously design, organic or inorganic (or both). It seems quite likely that the minds of such successors will have a larger set of things they can imagine than our own.
It also seems likely that these cognitively superior children of ours will be here within the next century. Presumably we will go extinct soon after their arrival (like all good parents making way for their children). So, as one of our last acts on our way out the door, as we gaze up at our successors in open-mouthed wonder, we can simply ask our questions of them.
What I should have said is that AI is only a part of the larger question of limits. And, of course, death is the ultimate limit, or so it seems.
His argument seems to me that humans are equipped with formal structures of cognition that are perhaps evolutionarily based and that are therefore basically set in place and relatively fixed. I agree that we interpret the world though conceptual schemes , but my contention that these schemes are continually adapting and changing. their nature in response to feedback from the world, so there is not the disconnect between formal cogntive structures and world that Wolpert suggests needs to be overcome in order to see more of reality.
Quoting 180 Proof
The essay also sketches out for further research the limitations of human cognition, something we're so proud of, even naming ourselves homo sapiens & raises the question of our tools' (logic, language, math, etc.) capabilities vis-à-vis unraveling the mysteries of reality.
I've always wondered, since I found out about some parasitic worms that have devolved neurologically in the sense lost their brains (imagine that!), whether, in the domain of cognition, we're the first ones to arrive or the last ones to leave in a manner of speaking? "To be smart is stupid" says Wolpert and that's very Zen, oui monsieur?
Speaking for myself, I believe there's a threshold in mentation which once attained is like a gateway crime - beyond it everything conceivable can be conceived, everything that can be grokked can be, no limits. Whether we're approaching that point - a singularity in its own right - or whether it's already under our belt is unknown.
Quoting 180 Proof
Maybe ... I sense ... a presence! :snicker: I'm into what you'd call woo-woo. I hope you don't mind!
He calls human language and mathematics "cognitive prostheses" and identifies "mathematics as a special case of human language.". We no longer expect ordinary language to give an adequate description of the physical world. Should we expect mathematics to?
Quoting Joshs
If I understand him, he moves in a different direction:
He makes a distinction between the possibility of gaining knowledge based on what we can imagine and our inability to imagine what that knowledge might be.
Chomskys innate transformational grammar module might be a good source of comparison. If we believe that language is the basis of conceptualization, and a syntax-organizing device explains the formal grammatical basis of language, then it is reasonable to claim that the arbitrary limitations this device sets on how we conceptualize also limits what kinds of knowledge we are capable of conceiving.
We might instead reject the claim that the content of what humans can conceptualize is limited in a fixed way by any sort of innate formal organizing structure. We can agree that for any given individual , within any given culture, at any given time, the way that we conceptually organize reality constrains what and how we are capable of understanding. But we would have to add that there is no way around this, and we shouldnt think that it would be an advantage to try to find a way around it, as if our conceptual constraints are causing us to miss something out there we might otherwise be able to perceive.
This is the wrong way of thinking about the matter.
The same conceptual schemes that constrain what we can understand are what make it possible for us to be able to conceive at all. There are an indefinite number of ways of construing reality, and any given way of organizing the world we commit ourselves
to will off close off other ways of thinking. There is no way of conceiving reality that is not potentially open to us,
but the only way to know in the future what today we cannot even conceive of is by starting from our current models , and exploring ways of transforming them, of turning them on their head. This is how science, and all other aspects of culture , already evolve.
To paraphrase and correct Wolpert, we regularly become those beings for whom things are knowable, but not to us currently, because we are not capable of conceiving of that kind of knowledge in the first place (within our current schemes of conceptualization).
Its not a question of getting out of this hermeneutic circle, but of getting into it in the right way.
It may not be a matter of the limits of current schemes of conceptualization, but of all schemes of conceptualization that are available to us qua human beings.
This is not a question that we can give a definitive answer to, but my guess is that we will never know more than a sliver of reality.
I would say yes, if by being youre getting at some kind of metaphysical a priori. In other words, I wouldnt expect that if we were to discover a planet with its own intelligent life, its conceptualizing capabilities would be radically different than ours.
I am not so sure. I imagine it very well might be.
1. [math]\infty[/math] (Cantor)
2. 0 (Brahmagupta)
3. Zen koans, Nagarjuna's tetralemma (it is? No; it is not? No; it is and it is not? No; it neither is nor it is not? No)
4. Paradoxes i.e. contradictions (p &~p, logic bombs that wipe the slate clean, back to a tabula rasa/the mind of a child/the uncarved block)
5. Dimensions > 3
6. The unanswerables (vide Wikipedia, Noble Silence)
:zip:
[quote=Ludwig Wittgenstein]Whereof we cannot speak, thereof we must remain silent.[/quote]