http://mashable.com/2016/06/02/elon-musk-simulated-reality/#sdLXHm2_jsqB
2,657 thoughts on “Elon Musk Thinks Evolution is Bullshit.”
Leave a Reply
You must be logged in to post a comment.
The Skeptical Zone
"I beseech you, in the bowels of Christ, think it possible that you may be mistaken."
phoodoo,
Thank you for both referring to and exemplifying the Kruger-Dunning effect.
Kantian Naturalist,
So was evolution to blame? Your sophomoric banter is your white flag of surrender?
What does this mean they do better? What is the definition of better in this instance?
Better at avoiding the question, like KN, you mean?
As KN points out, that is priceless. 🙂
Good point, phoodoo. Better is only meaningful in the context of some performance criterion. Evolutionally speaking, this is usually expressed as number of offspring achieved.
I don’t think the Dunning-Kruger effect undermines my general thesis.
My general thesis is that, in general, accurate representations are functionally integrated into successful action, where the representations are action-guiding and affordance-detecting, and the action is functionally integrated into the goals that characterize an organism’s species-specific form of life.
According to this thesis, an animal could not achieve its species-specific goals if it did not have mostly accurate, action-guiding representations of the affordances that comprise its ecological niche.
The Dunning-Kruger effect shows that individuals who lack competence in some task are also unable to assess their lack of competence. This suggests that the cognitive processes underlying competence are recruited for metacognitive assessment of competence. If you’re not good at something and don’t know it, then you’re likely to think you’re better than you are. (Interestingly, there’s also an inverse effect, where the highly competent are likely to misjudge how difficult the task is for others who lack requisite training.)
The Dunning-Kruger effect is a nice example of cognitive bias. But it was never any part of the thesis I am defending that no organism lacks cognitive bias, whether perceptual biases or conceptual biases.
What would undermine my thesis is not cognitive bias, but rather an organism that was able to act successfully in its ecological niche even though its action-guiding representations systematically failed to detect the relevant affordances.
While one can imagine such cases, and hence such cases are logical possibilities, I myself don’t know of any such cases in the actual world.
And if there aren’t such cases, that’s sufficient to refute the EAAN. It does so because the EAAN is intended to show that naturalism + evolution is self-undermining. Remember: the idea is supposed to be that naturalist is not justified in taking her own cognitive faculties to be reliable, hence she should not endorse any judgments resulting from those faculties, and hence should not believe that naturalism is true.
But if the general thesis is true, then the naturalist has good reason to believe that successful action tends to flow from accurate representations of environmental affordances. Thus she has good reasons to believe that her cognitive abilities are in fact generally reliable in circumstances sufficiently similar to the ecological niche in which those abilities evolved.
The next question is whether those same cognitive abilities are also recruited in the process of doing science: conducting observations, collecting data, designing experiments, evaluating probabilities, and selecting hypotheses. What we do know about the cognitive science of scientific practices is that the answer is “yes”.
Now, that does not mean — and here’s where the “Darwin’s Doubt” as initially expressed in his famous letter to Asa Gray comes into play — that those cognitive abilities are also reliable for framing fully comprehensive metaphysical doctrines. But contra Plantinga, that’s no objection to the naturalist, because the whole point of naturalism is that metaphysical speculation will be completely unreliable if it is not constrained by empirical inquiry!
No, it is not a good point at all. It cut off what I wrote just where I was explaining what mammals and birds were better at, only to bleat another tedious question while ignoring the competent answers given. I had written:
It makes no sense to say it’s a good point when he’s only ignoring what was written in order to ask his uncomprehending, repetitive “questions.”
Glen Davidson
Makes more sense simply to ignore doodoo.
It’s a bet peeve of mine when people use good or better without any indication of for whom and in what respect.
Sure.
I also have a bit of a peeve with the word “intelligence” as any kind of useful descriptive when comparing across species.
No response to what I wrote, I see.
May as well ignore context, as phoodoo did, and make unsupported accusations.
Yeah, I wrote the truth, you just sneer.
Try to be better than phoodoo, at least.
Glen Davidson
I have more than a little distrust of someone who complains about someone using terms as they’re normally utilized.
It’s really a poor display on your part.
Glen Davidson
Not sure why you think I’m sneering. I agree about the biology.
I like to think of “intelligence” as “the ability to perceive, use, and ultimately create patterns”.
Granted, it’s hard to operationalize across species in a rigorous fashion, but that doesn’t make the concept useless or negate Glen’s point.
Probably because you complained about me not saying what “better” means when I did exactly that, while you let phoodoo’s mangling of the quote inform you of what I did. You just doubled down on your poor consideration of the matter.
So why would I think you were being any better in the second part?
Glen Davidson
Fair enough, Glen. Aiguy/ RDFish has made many posts at UD and some here about the abuse of the word. I’ve been persuaded by his arguments on the lack of any measurable property that might be described as intelligence.
No problem. As I said to Glen, there have been many discussions at UD involving aiguy/RDFish where he has pointed out the lack of any consensus scientific definition.
I’d say all but impossible, though wasn’t there some previous discussion here about demonstrating self-awareness in elephants using large unbreakable mirrors?
That cognitive level is proportional to amount of nervous tissue. No indeed. I agree on the biology.
Kantian Naturalist,
Right, and I agree that being intelligent means better able to perceive. And THIS is why you get it (and Plantinga-even though I don’t even need to bring him into it) wrong. The question is WHY do organisms evolve to be MORE intelligent, thus more able to perceive. As Glen so accidentally points out, he thinks there is a hierarchy of how well animals perform perception wise. And YET there is virtually NO difference as to how well these animals perform REPRODUCTIVELY! This is why evolution is useless at explaining intelligence. You don’t need intelligence if you believe in evolution, the empirical evidence shows that.
But this is of course a “human centric” definition of intelligence (As Alan seems to understand), because if we really were using an evolutionary explanation, we would say that whatever is surviving best is intelligent-perception be damned. And so probably beetles are the most intelligent animal on the planet.
But back to Plantinga, the only “empirical evidence” you have of your better perception is still only your better perception. It is not your better reproduction. And yet you still believe it, based on perception. So you might as well also believe in one’s perception of their religious beliefs. So we do have intelligence we trust, but its not because of evolution, its because we have intelligence. So if we trust this intelligence, we trust our religious beliefs. If we trust this intelligence because we think evolution gave us this, then we better trust Dunning-Kruger, and assume you are wrong.
Why you think you are the exception to Dunning-Kruger, believing in evolution as you do, is beyond reason. Beyond evolution actually.
Science can’t tell you that. It can only attempt to explain how things happen. “Why” questions are easy to ask and have no answers so far, other than the ones we make up for ourselves.
ETA quotation marks for clarity
Evolution is a very neat explanation for how life has developed over time on Earth. It fits the facts very well. There are no competing explanations that answer “how”. It really doesn’t equate to religious belief.
I don’t agree with that. Rather I think being more intelligent means being better able to act in a benficial way on those perceptions, depending on the circumstances.
I think one of the problems with evolving higher intelligence is that the differences between related species are so small that a slightly higher intelligence is almost (as in practically) invisible to selection. So it is not that intelligence isn’t beneficial, it is that it isn’t beneficial enough to break out of the noise of random drift in most circumstances. That it only really becomes strongly enough beneficial once the level of intelligence differences between species become huge.
In the cases where organisms evolved to become more intelligent I think it is actually mostly due to drift. As in chance. (I can already hear creationist heads exploding, thankfully there won’t be any shrapnel).
That would explain why there’s only really one species with a human level intelligence and why most lineages, to my knowledge, don’t show any, or at best very little tendency towards greater intelligence even over long periods of geological time. For us homo sapiens, now that we HAVE reached this level of intelligence, it has finally become strongly beneficial. We have passed a threshold so to speak.
Could be sexual selection. Smarter men get more women. smarter defined as whatever it takes. Only recently has academics existed.
Isn’t there a Y chromosome that is shared by tens of millions of asians?
I see the distinction you’re making, but I’d urge that we not get caught up in the use of “how” and “why” as lexical markers of that distinction.
“Why do mammals give birth?” is a perfectly fine question, even if it could be phrased more precisely as “What selective conditions and developmental constraints led to the emergence of this function?”
The distinction that needs to be drawn is between testable and non-testable explanations, not between “how-questions” and “why-questions”.
phoodoo,
The point I was making is about the relation between successful action and accurate representation. Whatever you’re saying here about intelligence has nothing at all to do with what I am talking about. If my thesis is correct, it would still be correct if the most sophisticated animal on the planet were a fly or bee.
In other words, it’s a thesis about the nature of animal cognition — or more precisely, a thesis about animal cognition in those cognitive systems that involve representations (and I do not insist that all do).
There are many ways of making a living on this planet, and most of them do not involve intelligence. Being smart is a trade-off against other energetic, mechanical, and developmental constraints. But it is a good trick for occupying ecological niches that are vacant by larger and/or more numerous animals. Mammals evolved their intelligence as nocturnal insectivores who had to calibrate movement with sound and smell. Primates became more intelligent as diurnal frugivores who had to rely on subtle cues to locate high-energy, hard-to-find resources and determine when they were ripe. (This is why primates re-evolved color vision while most mammals lost it — you don’t need color vision at night!) Other mammals occupied other niches in which intelligence was not as important, though cetaceans found their own route to intelligence as social carnivores. Hominids became more intelligent still as diurnal omnivores who had to scrounge out a living as cooperative foragers in a savanna occupied by large predators and scavengers.
So it’s not that intelligence is always adaptive or that the growth of intelligence is the only or even more important trend in animal evolution. It’s that it’s a trend, under specific conditions, often adaptive but not the only way of making a living. Millions of species do perfectly well without much of it at all.
As usual with creationists and ID advocates, what they think is an objection to evolutionary theory reveals only that they don’t understand it and don’t want to understand it.
A second reply. I first responded before the ETA, though I saw that when I posted my reply.
First, my reaction to Putnam’s argument. I’ve always been puzzled by that argument. I’m puzzled, because it seems obviously absurd. I’m puzzled that other philosophers have not pointed out how absurd it is. A natural language is not the kind of thing to which one could apply Lowenheim-Skolem.
George Lakoff took Putnam’s argument as a reduction of Chomsky’s linguistics. That’s at least plausible to me, since I saw Chomsky’s linguistics as absurd anyway. I suppose I could look at the argument as a reductio of the coherence theory of truth, though Putnam didn’t seem to be using it for that way.
So back to rings of continuous functions, which you somehow connect to that. I suppose I should have forseen that. However, I am not really using it in anything like the same way. I do not assume that the brain is doing any fancy mathematics at all. I was referring to the functional analysis only as a hint at how we might explain the emergence that I was calling on.
I doubt that there is anything like a ring of continuous functions involved in brain operations. To the extent that the brain could be said to be using continuous functions, I take those to be something like piecewise linear functions (small bits of neural processing stitched together).
If f, g, h are continuous real valued functions, then a relation such as F(f,g,h)=0 might be called a structural relationship between those functions. A ring structure is one kind of structural relation. However, it looks to me as if the alternative derivation of the functional analysis result can be modified to show that the structure of a topological space is implicit in the structural relations of the continous functions over that space. So we should be able to manage with simple structural relations rather than a ring structure. And it looks to me as if Hebbian learning would create those structural relations, doing what I would consider to be a program of calibration of the measuring/categorization that is being done by the perceptual system. But the neural system does not need to know the mathematics. It should just work. So, as said above, the reference to Gillman & Jerison is only for the sake of our understanding how the emergence would work.
Yes, we certainly agree about that! One sees this problematic conflation not just in Putnam but also in Quine, Carnap, and really going all the way back to Descartes’ attempt to do for metaphysics what he did for mathematics.
KN, to Alan:
Alan,
It’s the same error you were making back in January:
My response:
Good comment, but I thought it best to correct this. Many primates (old world monkeys and apes, some new world monkeys, and I’m not sure if there are others) re-evolved trichromatic vision (gaining red cones), while other mammals mainly have dichromatic vision.
Glen Davidson
keiths,
Right. I do think that Alan is making the right kind of point, if you think about this in terms of causal explanations and rational explanations.
If I want to know why do didn’t show up when you said you would, I might be asking for a rational explanation — for what reason you didn’t keep your promise. But if you explain that you caught got in traffic or there was an accident, you’re be offering a causal explanation for why worldly conditions didn’t permit you to fulfill your promise.
What Alan seems to be saying here is that there aren’t any testable or verifiable rational explanations for biological and cosmological phenomena.
All the verifiable explanations we have for biological and cosmological phenomena are causal explanations, and all the verifiable rational explanations that we have are limited to ordinary interactions and the social sciences.
If that’s Alan’s point, I would entirely agree, but that has nothing to do with the use of “how” and “why” to mark the distinction.
Another late response.
Okay. Then there’s been some miscommunication. I was using “rules” where I should have used “mechanisms”. And I see mechanisms as prior to norms. Roughly, norms are the statistical effect of multiple mechanisms working on something similar.
In your book, you describle CI Lewis as assuming a semantic given. I’m pretty sure that I am not doing that. But now I’m wondering if norms are your assumed given. To me, norms are to be explained rather than to be part of an explanation.
That’s a weird way of putting things. I don’t see that I am suggesting classifying internal states. Rather, I am using internal states to categorize an external world with which we interact. And I see categorization as prior to sensory information. I see sensory information as information about categories, so we cannot have sensory information without first doing some categorization.
This is our major disagreement. I don’t see that there could be “real patterns”. That is to say, we find patterns in our representations, rather than directly in reality. I’m not sure what “pattern” could mean, that is independent of our representations and concepts.
OK.
Whether we are explaining something in terms of norms, or norms in terms of something else, depends on context. Both are legitimate strategies. I explain justification and assertion in terms of discursive norms, I explain the emergence of discursive norms in terms of niche construction of obligate cooperative foraging, and I explain teleological norms in terms of development and evolution.
I agree that our perceptual experience of the world has implicit categorization built into it. Given what little I know right now, my inclination would be to explain this in terms of how predictive processing implements, in our subpersonal cognitive machinery, anticipations and expectations of movement-correlated sensations, and sensations convey the prediction errors to the prediction-generating cognitive maps or models of the actual environment and its possible states.
I agree that patterns in our experience are generated from sensorimotor abilities, but what makes a pattern a “real pattern” is whether it also picks up on information in the world that is there anyway. If there weren’t any real patterns at all, then nothing about the world would determine whether our embodied coping is successful or unsuccessful.
I only meant to point out that I had no objections to sophisticated mathematical results being used in philosophical arguments, as long as they were correctly understood and then linked to the argument.
No further analogy was meant.
Thanks for the further description of your views.
I understand norms here to be needed to specify when this group of mechanisms is functioning correctly. How do see statistics being able to separate the correct from the incorrect?
I would have thought that is what philosophers of physics do when they read it, but not what physicists are doing when they read it. But then again, I cannot say I read much theoretical physics myself, at least not anything beyond the sophomore level.
In any event, I’m intrigued by your phrase “the structure of reality” since I read previous replies from you to KN as denying any such thing.
Based on your philosophy of math, I am going to guess you take a fictionalist view to this structure, ie what you meant was something like “according the narrative of this theory, the structure of reality”. Am I close?
Don’t find this convincing. There’s a huge cost in resources and power consumption in having a large brain. In humans, birth is a more difficult and risky event. There had to have been evolutionary advantages to larger brain size. I’m dubious about drift at the best of times, but for the trend to larger brains? Homo erectus had a pretty big brain and the hyoid bone indicates they could have had the ability for complex vocalisation. Modern Homo were around for maybe 70 to 80 thousand years prior to civilisation. As Petrushka mentions, there’s a good argument to be made for sexual selection.
The trouble is it works very well for me. the child’s “why” always leads to infinite regress and is ultimately unanswerable.
I think you are making my case for me. 🙂
Well, OK. My point was that science is not a substitute for religious certainty, belief or authority. I’m not going as far as Gould but…
ETA plagued with slow internet this evening making commenting most frustrating. Taking a break.
I think it’s the only plausible argument for selection as the cause of big brains.
You obviously don’t need a big brain to survive “in the wild”. But it might help a guy pick up girls. Or stifle the male competition. Big brains as somehow analogous to antlers. It definitely happens in agrarian societies. It might have happened because smarter people were better at providing food, particularly in hard times.
Considering the cost (mostly in the birthing process) I’d argue against drift.
Putnam’s target was how a metaphysical realist can explain reference and intentionality of both mental and language symbols. MR for Putnam does include Correspondence Truth, so that was part of his target.
From SEP
The argument purports to show that the Representation Problem—to explain how our mental symbols and words get hooked up to mind-independent objects and how our sentences and thoughts target mind-independent states of affairs—is insoluble.
The author of the SEP article, who is a metaphysical realist, wrote a book which characterized Putnam’s MTA as one of the most difficult for MRs to deal with.
People talk about the structure of reality, which is what I meant there.
My way of describing it, would be “how we structure reality for our own use.”
In my view, there isn’t information in the world (unless we put it there). There are only signals.
However, to be fair — 20 years ago, I would have talked in about the same way as you are talking. I’ve since changed my mind about what we should mean by “information.”
Perhaps my being a mathematician colors what I mean by “pattern”.
Every time that I press the door button, I hear the chime. But that’s not a pattern unless I repeatedly press the button. It is a causal relation. So maybe you are using “pattern” for what I would consider a causal relation.
And a side note. When my mother used the word “pattern”, she was usually talking about something she would use in sewing (something like a sewing template).
KN,
If his point were simply that there are “how” questions and “why” questions, requiring what you’re calling causal and rational answers, respectively, then yes, he’d be right. But that isn’t what he’s saying.
He’s being hyperliteral. For him, the actual word used is the determining factor. He therefore takes phoodoo’s question…
…as necessarily a “why” question and criticizes phoodoo on that basis.
He did the same thing with Schueler. He recognized that Schueler’s implied query was a “how” question, leading him to insist hyperliterally that Schueler should have used “how” in place of “why”:
That’s ridiculous, of course. Schueler’s statement is perfectly good English, in no need of Alan’s proposed substitution.
Only if one is a metaphysical realist about causal powers (as indeed I am) that give us a world with genuine causal (and modal) structure to which our strategies of embodied coping (including scientific practices) are answerable. I think that’s rather more realism than you’re comfortable with, no?
I don’t have a problem with causation, though I’m inclined to take it as derived from what we are able to cause.
You perhaps misunderstand my disagreements with realism.
My view, roughly is this: We should not ask what exists. Rather, of the uncountable infinity of things that exist, we should ask which are important enough for us to single out and name.
I see some arbitrary choices made in our categories and our scientific laws. If another group of humans made different choices, they would have different categories and laws. But maybe a translation would be possible — or an extended system of categories and laws would be possible that could allow some translation. However, if aliens from Andromeda made their choices of categories and laws, those might not be inter-translatable at all because they might depend on a very different biology.
Rumraket:
Alan:
I agree. Human brains comprise 2% of our body weight but burn 20% of our energy. Natural selection wouldn’t have tolerated such a massive cost during hominid evolution if there hadn’t been a compensatory benefit.
KN:
Alan:
It doesn’t work well at all. You wrote to phoodoo:
Now look at the following questions, all of which qualify as “why” questions according to your excessively literal criterion, and ask yourself: can science answer these questions?
1. Why is the moon pockmarked with craters?
2. Why are polar bears white rather than black?
3. Why does potassium explode when placed in water?
4. Why can’t I throw a rock into orbit?
You get the point… I hope.
Alan:
If the possibility of such a regress were disqualifying, then “how” questions would also get the axe. Think about it.
KN:
Alan:
Quite the opposite. He’s showing that “Why do mammals give birth?”, which is a “why” question by your criterion and thus supposedly unanswerable by science, is equivalent to a question that can in fact be answered by science.
It refutes your claim…
…just as my comment above does.
KN:
I disagree with you on this point. Both are useful distinctions (provided the latter is taken non-literally, contra Alan) and they are orthogonal.
Neil,
Signals carry information. See the “sign” in “signal”?
There’s some evidence against this suggestion. For one thing, behaviorally modern humans seem to come on the scene by 40,000 years before present, whereas agriculture goes back (I think) no more than 12,000 years. For another, there’s a good deal of evidence from comparative cognitive psychology that one of the things we’re much better at than other apes is cooperating.
Humans are different from other apes in being obligate cooperative foragers. In chimps, the band will forage independently and come together for socializing, playing, mating, etc. But in human societies, from small hunter-gatherer tribes to large WEIRD (Western, Educated, Industrialized, Rich, and Democratic) civilizations, we see many different people occupying many different social roles that work together for the provisioning of the group as a whole. (This anthropological fact may be somewhat obscured by the value of competition in capitalist societies.)
There’s some evidence that obligate cooperative foraging is an ecological niche unique to hominids (though some cetaceans seem to have something similar), and it played a crucial role in how hominids were able to sustain themselves in the Pleistocene savannas. My best guess is that the ecological value of having a big brain lies in its role in promoting in-group cooperation, not better in-group competition. (I say that because there’s still going to be plenty of inter-group hostility, sometimes negotiated by trade and by marriage and sometimes by warfare.)
Some chimpanzees coordinate their hunting activities to an amazing degree. The chimpanzees of the Taï Forest in Côte d’Ivoire hunt red colobus monkeys in coordinated groups, playing roles such “driver,” “blocker,” “chaser,” and “ambusher” that can require 20 or more years of experience to master. Meat is distributed to all participants in the hunt, with shares reflecting the importance and difficulty of each role.
See: Boesch, C. (2002). Cooperative hunting roles among taï chimpanzees. Human Nature 13: 27-46.
I didn’t understand that point of Neil’s either. I hope he will explain.