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."
keiths:
Bruce:
True recognition, of the semantic engine variety, requires norms, though not necessarily public ones. ‘Sorta’ recognition doesn’t require norms at all — just syntax (i.e. physics). Dennett’s point is that the way Jones sorta recognizes a horse/schmorse is akin to the way a two-bitser sorta recognizes a quarter/quarter-balboa. Norms play no causal role.
Right, with the proviso that different users, or even the same users under different circumstances, can have different requirements The two-bitser can serve both as a quarter detector and a quarter balboa detector, and there is no fact of the matter about what is the real meaning of its internal state.
keiths:
Again, I would say that there is no fact of the matter regarding who those names refer to, just as there is no fact of the matter concerning what the two-bitser’s state means.
Glancing at “explicit criticism” of the Wikipedia entry on swampman, I see my point has been made before.
Your point is obvious, which is why I specified my scenario the way I did:
Your objection applies only to your own example, not to mine.
Your version is logically impossible in my view, using “identical” as meaning identical.
PS The more important criticism (made by Dennett apparently) is when a thought experiment strays too far from reality into imagination, it is not very useful.
ETA Just to show willing (and having read Dennett’s acknowledgement of his debt to Rorty), I’ve downloaded Sweet Dreams.
I disagree in the strongest possible terms. If someone wants to assert that X can be reduced to Y, then the burden of proof is one the person making that assertion to provide evidence for it. I am not saying that X is irreducible to Y. I am saying that there’s no evidence to support the assertion.
Let’s put it this way: there are, of course, some good examples of successful intertheoretic reduction in the history of science — though one must immediately note as well that scientific revolutions are often replacements of previous theories with new ones, and how much of the older theory is reduced is a difficult question.
Ernst Nagel argued that successful reduction involves explaining how higher-level concepts (or concepts with a more limited domain of application) can be effectively treated as special cases of lower-level concepts (or concepts with a more generic domain of application). For example, Nagel argued that we can reduce thermodynamics to statistical mechanics by treating pressure and temperature as special cases of force and mean kinetic energy
But at the same time, it’s still controversial whether Mendelian genetics can be reduced to molecular genetics (see here).
I think that if we’re not even sure what reduction means and whether nor not classical genetics can be reduced to molecular genetics, it’s going to be really hard to say that the sciences of cognition — ecology, psychology, and neuroscience — can be reduced to the theories of fundamental physics (quantum mechanics and general relativity).
Hah. You want to revise my example now?
Look, this isn’t really so complicated–at least where someone has conceded that he cares about himself more than about his double and someone else–can mean to refer to that person which was above cared more about and not be referring to that double.
I think this issue is being needlessly complicated by the “swamp” aspect, which involves, as Alan has noted, a bunch of sticky issues involving identity that have nothing at all to do with my (basically Putnam’s ‘twater’) case. In my example, there’s no need for “molecular identity”: the doubles must simply have the same superficial characteristics (i.e. look and act the same except, of course when being analyzed by chemists) and also, when created, have been the recipient of all and only their model’s thoughts, memories, “mental propensities,” etc.–perhaps have had them “uploaded.” (I make no comment on whether such an “uploading” even makes sense).
So let me put it this way. A twin of the city where you live was created one year ago. You get painlessly transported there in your sleep last night, and this morning you see someone you think is your good friend Richard at the same time as twin Keiths sees this person/thing. You both exclaim, “Hey, Rich!” simultaneously. YOU are referring your friend back in Sacramento. Tweith is referring to Twichard. He’s got the guy he thinks he’s referring to in front of him; you don’t. Now, if Twichard (who hasn’t noticed Tweith) turns to you and says “Hey, Keith, how you doing?.” He’s referring to the wrong guy, too.
I don’t say the twins don’t speak English, and I don’t know if Bruce would either, although you keep saying that–I’d prefer to hear it from him. But it seems to me that you are clearly meaning different things by “Keith” and “Richard” –because after a year, whatever he’d meant before, Swamp Keiths is now talking about that thing–his swamp buddy.
It’s hard to make this example fit into a molecular identity picture because the twins (like the twater there) are specifically constructed in the example NOT to be structurally identical. If you want that to be part of the example, I don’t think we can have both Keith and Swamp Keith present at the same moment (or it’s more complicated story, anyhow). But we can still note that at the moment that Swamp Keith and Swamp Richard plop into existence, see each other, and say “Hey Keith” and “Hey Richard” simultaneously, if, at the same moment, you and Richard visiting a nearly identical swamp and say the same things, it’s not nuts to suggest that two of you (the real guys) are right and two of you (the swamp guys) are wrong.
I don’t opine on how these interpretations affect any thesis of Dennett’s (or his ‘privileged access” non-sequitor) that you quoted above.
Is there is a reality which makes your interpretation wrong or which cannot be captured by a backwards looking interpretation.
Standard example: did Caesar sneeze in the night of his sleep before he was assassinated?
Yes, I agree that is an important point.
The complete randomness is key. It is not random in the sense of a radioactive disintegration, because in that case we have a probability distribution. It is a random in the sense of being a completely unique event, which is not subject to scientific law. So it is very questionable whether we can use science to analyse it, eg to predict swampman’s behavior using its stipulated structure and then applying scientific laws.
If one is trying to provide a scientific account of meaning and understanding, perhaps that is enough to reject the relevance of swampman arguments. That’s Dennett’s view.
But I still think the causal history issue to be used when determining norms is also relevant and helpful
Another standard example for history and norms: what if a semi-mad scientist operates on you in your sleep to create a wound indistinguishable from a mosquito bite. Does that make it a mosquito bite?
Note that there is a causal history here so I would be comfortable with saying things like it will itch regardless of whether it is correctly called a mosquito bite.
So are there ever facts about meaning and understanding?
If so, under what circumstances?
ETA: Have we changed the subject from discussing the claim that SM does understand to discussing the claim that there are no facts of the matter about whether it (or you or me) understands?
Seems that way to me, but maybe I am missing something in the history of the thread.
I don’t know. If you have evidence, that will be evidence that exists in the present.
I think you are missing the point. Your view of Swampman throws serious doubt on all of history. We might have all popped into existence yesterday, with memories intact. On your view of Swampman, none of us would have any understanding.
I’m simply taking a view of history which is not susceptible to that kind of skepticism.
I guess it depends on what “understanding or speaking English” means. I am taking the meaning of that phrase to incorporate the references to the world that members of the English-speaking linguistic community share.
For me, in Putnam’s twin earth, the twin speaks Twenglish because his or her references for eg natural kinds are different.
Or if the twin does social media, then the twin posts TWenglish on TwTwitter.
I’m not sure I understand your example: are all of the twins on our earth? I am not sure what I would say in that case. Part of my hesitation relates to linguistic communities and linguistic division of labor within them. How different do the communities and their context have to be before we can say they are speaking different languages? I’m not sure where to draw the line. But I do think Putnam’s twin earth is beyond that line regardless of where exactly it is.
So extending that point about linguistic communities: if an entity has no causal relation to any community, can that entity at the moment of its creation use language to refer or understand references in English as used by Keith or me ETA: or twMe or twKeith?. I say no.
ETA: BTW: these points miss the issue of the role of mental representations in understanding a language. Both me and twMe have the evolutionary history to justify attributing norms to the representations created by our cognitive mechanisms. Swampman lacks that history. So that is another reason for doubting its ability to use language. However, as I recognized yesterday, that does present problems for my point that eventually it would understand the language spoken by its adopted linguistic community. I don’t have an answer for that apparent contradiction in my views.
I do agree if we assume we are all random beings just created, then there are no norms for meaning and so yes our behaviour in the intitial moments is not correctly called understanding language. That situation will change over time assuming we remain in existence.
We are not members of the species homo sapiens either, since the definition of species depends on biological history.
Got a bigger bullet for me to bite?
My comment was on your version of anti-realism. If you were making a point about my claim for the need for causal history and context to provide norms meaning in that original post of yours about history being a matter of interpretation, you are right, I missed it.
Also, it is not my swampman, it’s Davidson’s.
Alan,
I await your argument in support of that claim.
So far you’ve only offered an argument against your own Captain Kirk example.
Alan,
What is “too far”, and how does one make that determination?
And if Swampman is “too far”, then precisely how does he lead us astray?
atomically identical twins having different histories is voodoo stuff. Magic. God making a stone too heavy for him to lift.
It might reveal some paradoxical use of language, and if that is the point, that’s okay.
The interesting question for me is whether we can understand intentionality in terms of some consilience across ecology, evolutionary theory, and cognitive neuroscience, and — for discursive intentionality — linguistics and sociology. But I see no promise in the idea that any or all of those sciences can be reduced to the theories of fundamental physics.
I did see it, and found it very interesting! I’ll go back and look at it more carefully. I think I’m still struggling with the fact that predictive coding theory and embodied/embedded approaches are not as compatible as Andy Clark assumes. Presently I’m reading Gibson’s Ecological Approach to Visual Perception to get the original version of the story. Then I’ll go back over Noe and Clark.
KN,
Your reasoning still seems backward to me. There are plenty of phenomena that, as far as we can tell, are completely explained by physics. If you can’t provide a single counterexample of a phenomenon that is known to be irreducible to physics, then why shouldn’t we conclude (provisionally, of course) that everything does reduce to physics? That’s how science works.
Shouldn’t there at least be some evidence against that hypothesis before we reject it?
Also, the question isn’t whether we can actually succeed in reducing a phenomenon to physics, but rather whether the reduction is possible in principle. It would be silly to demand that stagflation actually be reduced to physics before deciding that a reduction of economic phenomena is possible, in principle.
Finally, the best evidence in favor of the reducibility thesis is not that we have so many examples of phenomena that can be explained in terms of physics, but rather that our best physical theories are causally complete (modulo quantum indeterminism).
If physics is causally complete, that leaves no room for downward causation, and the special sciences are just convenient redescriptions of what is happening at the level of physics.
petrushka,
No, there’s nothing logically inconsistent about it.
I tend to think of reduction as analogous to diagnostics.
Perhaps because I do mostly diagnostics for a living, and quite a bit at home.
When the refrigerator ices up, you don’t go to a physicist for an explanation. You look at higher level modules. If you are an engineer trying to make a product more reliable, you look at modules within modules. If the failure is in an electronic part, you might need a chemist. Other parts might require someone with knowledge of physics.
One interesting thing about malfunction of human brains is that people have attempted to diagnose troubles at the level of history, at the level of economics, at the level of sociology, at the level of psychology, at the level of chemistry (currently in vogue) and at the level of genomics and evolution.
Identicalness prevails at the level of particles. At the level of human-size entities, identicalness is unachievable. If you claim entity A is identical in every respect to entity B, then they must exactly match in space and time. Any entity in a different space or time will have different experiences, even if sharing a locked cell.
When you start to wonder “is it just me?”
Been skimming through Dennett’s Sweet Dreams. Good stuff!
walto,
No, I wanted to know where you got the strange idea that if walto means Keith when he says ‘Keith’, then Swamp walto means ‘Swamp Keith’ when he says ‘Keith’:
Having read your “twin city” scenario, I can see where you went wrong. You were confusing two separate thought experiments: Swampman and Twin Earth. In your mind, “Swampville” was a local version of Twin Earth, containing a Swamp walto, a Swamp Bruce, and a Swamp Keith.
But of course the fact that lightning zaps a Swamp walto into existence doesn’t mean that it does the same for everyone else, including Bruce, Keith, and Barack Obama.
That’s why I commented:
Gotta keep those thought experiments straight, walto. If you jumble them up, they’re not going to be very useful.
On a related note, I’m starting to delve more deeply into Friston, including a at least a higher level understanding of his math.
He seems to think PP is a form of DST, I believe. Which makes sense, since DST is simply coupled differential equations. As I recall, the blog post makes a similar point.
I’ve also watched a Friston video describing how Markov Blankets fit into his formulation of PP. Based on my current understanding, a Markov Blanket is not what I first thought it might be when applied within the math of PP. I had taken blanket as similar to veil as in Veil of Perception; that is, I thought the blanket bit implied a Cartesian approach which includes skepticism about whether our perception/action tracks the external world.
But as I understand the math, that is not the case. The blanket states are the sensory and action states and the internal states are brain states a (internal according to the math, not embodiment, since embodiment includes the sensory and active states). PP provides a math model of how the internal, sensory, action, and world states are linked. That math says that as long as the internal states are such that they are maintained far from thermodynamic equilibrium then those internal states must represent (in an information theory sense at least) the external states.
Once you get a bit further in your research with Gibson and Noe, I’d be interested in why you think PP is not compatible with embodiment.
You’d probably also like his Kinds of Minds.
Yes, precisely.
If you mean ALL of them, that seems to me obviously too strong. And, actually, I’m not sure I’d agree that you need ANY of them. There’s an awful lot of English left, I think, or at least enough so that Twenglish seems clearly derivative.
His sneeze, if it happened, does not exist in the present. I understand you to be saying that whether or not that sneeze happened depends on our current evidence (or lack thereof) and how we choose to interpret it.
Haha. I wonder if swamp keiths and tweiths are as careful to avoid answering questions and entering into the specific issues being discussed as our actual keiths is. I guess they’d HAVE to be to be at all similar to him…..
Can you explain that? I don’t see why that has to be the case. I’d think twin you could be a robot.
BruceS,
In my view anti-realism (or idealism of Neil’s type) requires constant confusing of ratio essendi with ratio cognoscendi–or as people here like to say–territories with maps of them. That’s basically the entire game.
The references of the rigid designators for me and TwMe are different because the references are relative to whether we choose our earth or twEarth as the actual world for determining that reference. For me, that is enough to say that I and twI don’t understand the same language.
I agree there is a lot more to a language than its the reference of its rigid designators. So my condition is sufficient, not necessary.
I’m not sure why you would say TwEnglish is derivative. It and English are independent languages with their own separate (cultural) evolutions and separate language communities.
Suppose my twin and I got swapped in our sleep. When we wake up on twEarth, someone asks me for a glass of water. Do I understand him? If understanding involves external (as well as internal content), I think the answer is no.
By the way, I may have a JTB that I do understand the question, so it is a Gettier situation, I think.
It’s interesting how different your intuitions about this are from mine. I’d say they were both English. Twenglish seems to me similar to keiths’ so-called “private language” where he calls cows “shmows”. The particular items being referred to (or mistakenly missed) don’t seem essential to me. I’d say it’s still English.
I guess I’d say that “understand” has to be unpacked a little here. It may be that you two don’t mean the same thing by “water” (and perhaps not by “glass” either)–but “understand” is a little squishy. You understand that a clear thirst-quenching, odor-less, tasteless, harmless, liquid is wanted–stuff that’s needed to survive, but not to be squirted or dumped on the person’s head, but to be imbibed; etc. That ain’t nothing.
If it was a designed robot then causation is at play. So it is not like SM.
Does it learn language like a child? Then it understands it.
Or is English hard coded somehow? Then we still might be able to make an argument that it understands English under the right causal situation describing how that hard coding was established. But if we don’t include interaction with our language community and its world, then we are back to whether the Chinese room as originally formulated by Searle is a good test for understanding. I think that understanding can only come about and should only be attributed when the robot has interacted with a language community in that communities world.
What about the cognitive mechanism? Here we have to start by relying on the norms of the designer being the same as ours to provide the robot’s norms. Now maybe after a time we find the robot will be well functioning in the real world. So that might be reason to transfer the norm-source to the robot. (That is Tye’s answer to the issue). Further, if we have a lot of experience with similar robots, we might be willing say it will be well-functioning without seeing that behavior.
In the above, I have ignored whether the robot was created here or on twEarth. That does not affect things as far as I can see, due to the period of interaction with a language community that I insist on before understanding occurs or should be attributed.
Anyway, at this point I would say we are pushing the boundaries of both philosophy and science. The science ones tend to push back harder, I have found. Newton has more force there, I suppose.
That’s what I thought it was, too.
The cognitive states must represent the sensory and action states, you mean?
I think the worry here is that the embodied mind people want to say that states of bodies and of environments can be constitutive of at least some cognitive states, whereas the PP people want to say that cognition is solely constituted by skull-bound processes that are coupled to bodies and environments. In other words, there’s a metaphysical issue here about how we distinguish constitution from coupling.
I take those sort of descriptive elements as the types of things Putnam might have been including in his depiction of meaning as involving more than just reference.
But reference is still a constituent of meaning and so changing reference changes meaning. And I say if we don’t mean the same things by our words, then we don’t have understanding.
By the way, one issue with that descriptive approach is a lot of the words used in those sorts of descriptions may themselves have twin earth type stories for them and so could be considered to have external content. It is a worry, not something I have the inclination to make more specific for now.
ETA: Maybe you meant for me to focus on understanding versus twUnderstanding? I guess if we cannot use “understanding” in some sense that applies to me, twMe, and SwampMe, we are up the creek and stuck in the incommensurability swamp without a paddle.
ETA: As partial compensation for being stuck in such a swamp, we may meet some interesting entities there.
Whether or not we say that the sneeze happened depends on current evidence.
It’s just that I see no way of distinguishing between “it happened” and “according to the evidence, we should say that it happened”. To distinguish between those would seem to require something comparable to a belief in ghosts.
In the case of a past which we cannot revisit, we only have a map. Any territory can exist only in our imagination.
walto,
I’m happy to respond to your points later this afternoon, when I have time, but would it kill you to admit your mistake, just this once? You got the Swampman scenario mixed up with Twin Earth. Embarrassing, yes, but not the end of the world.
Sure, but evidence of X and X aren’t the same things.
Exactly what one might ask you (only it’d be truthful instead of bullshit).
I think it’s important that “understanding” not be a Putnamian “natural kind” word.
That’s pitiful, walto.
Your Swampman/Twin Earth mixup is as obvious as it was earlier in the thread when you confused “odds that” with “odds against”.
You still haven’t acknowledged either mistake.
Alan,
Practical achievability is not the issue. Logical possibility is. This is a thought experiment, remember?
I stipulated that they are physically identical, atom-for-atom. There are obviously other differences between them. One was created by a lightning bolt, fercrissakes.
No, it isn’t logically impossible for two identical entities to have the same experiences despite despite being separated in spacetime. All that matters is that every causally relevant factor is the same between the two locations.
In any case — and this is the point you’re still not getting — I anticipated and defused your objection by specifying my scenario this way:
If the exact time and place are causally relevant, that can easily be accommodated. Just stipulate two possible worlds, one in which Swampman plays the role, the other in which non-Swampman is substituted for Swampman.
walto,
You’re not getting it. Put Keith and Swamp Keith into that same scenario — being asked whether it makes a difference to them whether they, or their counterpart, is shot — and both will say it does.
What else would you expect? They’re physically identical, after all.
walto,
This is just Jones and the horse/schmorse redux. You are imagining some kind of metaphysical tether between a brain state and its “true” referent. That’s as silly as postulating a metaphysical tether between the two-bitser’s state and either quarters, if you’re in the US, or quarter balboas, if you are in Panama.
You even say that the metaphysical tether can dissolve and reattach itself to another referent:
How did you determine that a year was sufficient? Does it depend on how often Swamp Keith thinks about Rich/Swamp Rich? How often he sees Rich/Swamp Rich? Did you run experiments to determine the metaphysical crossover point?
The notion of a metaphysical tether that gradually comes to tie ‘Rich’ to Swamp Rich is absurd. It makes as little sense as postulating a metaphysical tether between the two-bitser’s state and quarters, gradually dissolving and reattaching to quarter balboas after the machine is moved to Panama.
Alan:
Bruce:
The laws of physics don’t “care” about how you get to a particular physical state. They just say that once you’re in that particular state, future states are determined (modulo quantum indeterminism) by the interactions with the environment. The history doesn’t matter.
And even if it did actually matter that Swampman was produced randomly, that could easily be addressed. Just stipulate a deterministic process that generates Swampman as an incidental byproduct. The randomness is now gone, but you have the same Swampman as before.
Bruce,
No, because in that case the history is actually specified by the term “mosquito bite”.
The phrase “understands English”, by contrast, is a description of a capability, not a history. You obviously disagree, but in so insisting, you do extreme violence to the widely accepted meaning of the phrase.
Consider this scenario:
Swampman is zapped into existence. You immediately say to him, “Tell us about Lassie.” He launches into a long disquisition on the TV show and its titular character. He explains that she is a dog, a collie to be specific, and a loyal companion to Timmy. He may even mention, as is incorrectly but widely believed, that she fetches help when Timmy falls down a well.
All of this indicates that he understands English. Not only does his behavior so indicate, but his subjective experiences do, too. For example, when he hears the name Lassie, he forms a mental picture of her, complete with long hair and snout.
Yet according to you, Swampman doesn’t understand the question, he doesn’t understand English, and he doesn’t the detailed answer that he gives. He doesn’t know what “Lassie” refers to, and he isn’t able to picture her in his mind.
Why? Simply because he lacks the history that you claim is a prerequisite.
It’s a bizarre position, and it isn’t what people mean when they say that someone “understands English”.
Thanks, I’ll check it out. Does it have a chapter on Rorty? 🙂