Hi bitbutter, the distinction I want to draw is one of norm-governed behavior. There is a hugh difference between my dot matrix printer not printing and me refusing to talk with my wife after a fight. The difference is you can tell me that I'm -wrong-, I _should_ talk with my wife. There is an important sense in which you can't say that to the dot matrix printer.
Hi bitbutter, I would agree that the traffic control system makes choices only if, if it were to make the _wrong_ choice and somebody went to jail, then we'd take it to court and charge it with murder :-) I don't think that mere differential response to stimuli, no matter how complex, can be called a choice unless it is norm governed behavior.
Hi bitbutter, w.r.t. somebody standing at a crossroads... I haven't made the vid on this yet, but there's nothing to say that the reason which causes the choice ("you've got to go some direction!") is _deterministically_ related to the action. e.g. we still say smoking causes cancer even though not everybody who smokes gets cancer.
I'm willing to charge my PC with anything from fraud to arson to high treason at times. :D But then again, we do repairs and software updates; why is it that those COULDN'T be seen as reactions to violating norms (and better reactions than incarceration or death penalty, I might add)? Sure, they're not *perceived* as judgemental or like, but what is the reason they couldn't be elevated to such status?
I.e. the reason we perceive that machines don't make choices is because we don't perceive normative behaviour. But the reason for not perceiving normative behaviour can't be, then, that they don't make choices. So what exactly is that important sense in which you differ from a dot matrix printer? I'm still slightly not understanding your stance.
"So what exactly is that important sense in which you differ from a dot matrix printer?"
(For me the difference is a matter of degree: i am more black-boxish, i maintain state in a more complex way, and i'm more goal-ish than a printer. At a certain point of complexity it starts to be useful to use punishment on goalish, state-maintaining, black boxes to get the behaviour you want out of them).
Hi bitbutter and naphra, I think that we could agument a personal computer to the point where we could say it was making choices, but I don't think its merely the _degree_ of complexity. I'm trying to describe the _kind_ of complexity it would have to exhibit (cont)
(cont, to bitbutter and naphra2) I don't think we can attribute norms to the computer because doing so doesn't help us predict and controll it. You control somebody using a norm by _shaming_ them when they violate it. You can't do this to a computer. Until you can, it does you no good to attribute normative behavior to them.
But still, we CAN control and manipulate computers. What I'm asking is that why is it inappropriate to call that shaming? Yes, we're not used to _thinking_ that as shaming, but what exactly _makes_ it not shaming? To me, it isn't obvious at all that a computer can't feel ashamed, or at least, again, the reason to assert that they can't feel ashamed can't be that we can't shame them, for reasons of begging the question.
Hi naphra2, because we can only shame something that knows that it can be wrong. the computer can say to itself "gee, I really shouldn't let that program crash, that's against the rules"....... there are many ways we can control somebody--pay them money, beat them, shame them. These are distinct ways of control; we must draw the distinction in precise speech.
(wrt complexity) Yes, I can agree also with you Randy, there's probably a qualitative difference involved as well, not any measure of complexity will do here.
"Do you think that the ambien driven sleep driver drove by choice?"
Yes. _And_, i think that the sleeping driver is, in an important sense, not the same person as the waking driver. And a consequence of that is that punishing the waking driver for the sleeping driver's crimes would be unjust.
Hi naphra2, I don't think there are two entities; I think there is one entity which either has some modules turned on or some modules turned off. I'm trying to analyze the function of the modules which are turned off in the ambien drivers case. What is it about the difference which is so special???
Yes, and I'm willing to go along with that to an extent (also some of it is a matter of semantics), but I'm slightly uneasy with an idea of a "free will -bearing module". But I do see what you're getting at (I think).
Hi bitbutter, if there were two different entities, its merely a case of one being the subset of the other. they use the same neural circuts to, say drive, eat, open fridge doors, etc and they occupy the same spatio-temporal space. Its not two different entities, its one entity with something turned off. I'm interested in what has been turned off. That is what I'm focusing on.
"Its not two different entities, its one entity with something turned off. I'm interested in what has been turned off. That is what I'm focusing on."
I want to distinguish between entity (system?) and person. The driver and the waking ambien user are the same entity-with different modules active, but I consider the two different people (this way lies a foggy grey area: at what point does an amnesiac war criminal become a different person to the one who committed the crime? etc).
I think bitbutter's point _is_ anthromorphism and that it is intentionnal: what he wants to show is that a definition of choice would apply to hypercomplex computers and since computers exist in a whole scale of power and memory sizes and that since cannot really draw an arbitrary line anywhere, you find yourself considering anything as choice making.
maksiiiskam2: exactly so. Anthropomorphism is only a crime if the attribution of human characteristics to non-human things is inappropriate. It's yet to be shown that attributing choice making power to complex machines is inappropriate (and I think that's a hard case to make because people already use the word 'choice' in that way).
Hi bitbutter, suppose I hit somebody so hard that they die. Now suppose a dot matrix printer's spring pops out of the printer and hits somebody so hard that they die. I get charged with 1st degree murder. Does the dot matrix printer? Of course not. _that_ is the difference I'm trying to emphasize here.
Sure there's a big difference. But lets soup-up the dot matrix printer! Give it sensors, a memory, and a means of integrating the two to create new behaviour; If the dot matrix's was fitted with a 'brain' of such complexity that for practical purposes the machine's output could not be accurately predicted for any given input, we might have a machine that it would make sense to punish.
The Ambien sleepdriver case is a fascinating one, nevertheless I feel inclined to ask in the same vein, would you hold a person sleeping in the back seat culpable for a driver's offenses? i.e. who is to say there's only one distinct responsible choice-maker within one body?
Hi naphra2, sure, I'd put the asleep person in the "back seat" as it were. Your question is there more than one distinct responsible choice-maker within one body, which is a good question--basically, is there somebody else who is in the driver's seat here? There is some entity which is doing the driving, and that entity is distinct from the person which thee body belongs to--but is that entity a responsible, rational, agent? (cont)
(cont, to naphra2) here's how I would approach it: in order to expalin the behavior of the entity, do I have to describe the cases of its behavior in normative terms? For the ambien sleep driver, I don't think so; I think we can explain the behavior in purely descriptive terms, as the satisfaction of an urge. Compare it with, say, an amoeba, which hunts for food in a drop of water. We wouldn't say its reasoning--what does it reason with? Same goes for the ambien-entity.
Hm, I'm not sure if it's so obvious. Ok, perhaps the sleepdriver can be described as satisfying an urge (but then again, what urge does shifting gears satisfy? Rather, it's almost normative, as it's "good for the engine" to shift gears. Anyways.) But consider something like somniloquy where people at times narrate their dreams to vivid detail (Dion McGregor, say) or express opinions or emotions. It's not clear to me it could explained descriptively.
Or rather, maybe could be, but not to any better extent than explaining normal everyday waking-hour choices descriptively.
I guess I'm partially getting at what Az pointed out in his response, that the notion of will seems to me still left a little bit in the haze. Do we have A will, or a society of competing wills in our head, a la Minsky?
(Clips of Dion McGregor's somniloques seem to be available on Amazon if you want to have a peek, search Amazon for "Dion McGregor Dreams Again". It's really hilarious stuff, and I'm unable to conjure up any kind of straightforward descriptive explanation for some of it. ("No of course you can't play the radio!! Don't play that radio!! You're waking Mom!!"))
We are comfortable to say that complex machines make real choices. My contention is that humans belong to this group of complex machines. It's not clear what pareidolia has to do with this.
Excellent vid! Great points!
MikeSarno 4 years ago
'Anthropomorphism: The attribution of human motivation, characteristics, or behavior to inanimate objects, animals, or natural phenomena.'
Sure, I think that's exactly what Pinker is doing--with the aim of showing that human share characteristics with other systems.
If you think this kind of attribution is in appropriate, you'll have to make a case for that.
Do you agree that the traffic control system makes choices?
bitbutter 4 years ago
Hi bitbutter, the distinction I want to draw is one of norm-governed behavior. There is a hugh difference between my dot matrix printer not printing and me refusing to talk with my wife after a fight. The difference is you can tell me that I'm -wrong-, I _should_ talk with my wife. There is an important sense in which you can't say that to the dot matrix printer.
randyhelzerman 4 years ago
Hi bitbutter, I would agree that the traffic control system makes choices only if, if it were to make the _wrong_ choice and somebody went to jail, then we'd take it to court and charge it with murder :-) I don't think that mere differential response to stimuli, no matter how complex, can be called a choice unless it is norm governed behavior.
randyhelzerman 4 years ago
"I don't think that mere differential response to stimuli, no matter how complex, can be called a choice unless it is norm governed behavior."
So a person, stood at a crossroad with nothing to choose between the two directions, goes left. Are you saying he hasn't made a choice?
bitbutter 4 years ago
Hi bitbutter, w.r.t. somebody standing at a crossroads... I haven't made the vid on this yet, but there's nothing to say that the reason which causes the choice ("you've got to go some direction!") is _deterministically_ related to the action. e.g. we still say smoking causes cancer even though not everybody who smokes gets cancer.
randyhelzerman 4 years ago
I'm willing to charge my PC with anything from fraud to arson to high treason at times. :D But then again, we do repairs and software updates; why is it that those COULDN'T be seen as reactions to violating norms (and better reactions than incarceration or death penalty, I might add)? Sure, they're not *perceived* as judgemental or like, but what is the reason they couldn't be elevated to such status?
naphra2 4 years ago
I.e. the reason we perceive that machines don't make choices is because we don't perceive normative behaviour. But the reason for not perceiving normative behaviour can't be, then, that they don't make choices. So what exactly is that important sense in which you differ from a dot matrix printer? I'm still slightly not understanding your stance.
naphra2 4 years ago
"So what exactly is that important sense in which you differ from a dot matrix printer?"
(For me the difference is a matter of degree: i am more black-boxish, i maintain state in a more complex way, and i'm more goal-ish than a printer. At a certain point of complexity it starts to be useful to use punishment on goalish, state-maintaining, black boxes to get the behaviour you want out of them).
bitbutter 4 years ago
Yes, I'm inclined to agree.
naphra2 4 years ago
Hi bitbutter and naphra, I think that we could agument a personal computer to the point where we could say it was making choices, but I don't think its merely the _degree_ of complexity. I'm trying to describe the _kind_ of complexity it would have to exhibit (cont)
randyhelzerman 4 years ago
(cont, to bitbutter and naphra2) I don't think we can attribute norms to the computer because doing so doesn't help us predict and controll it. You control somebody using a norm by _shaming_ them when they violate it. You can't do this to a computer. Until you can, it does you no good to attribute normative behavior to them.
randyhelzerman 4 years ago
But still, we CAN control and manipulate computers. What I'm asking is that why is it inappropriate to call that shaming? Yes, we're not used to _thinking_ that as shaming, but what exactly _makes_ it not shaming? To me, it isn't obvious at all that a computer can't feel ashamed, or at least, again, the reason to assert that they can't feel ashamed can't be that we can't shame them, for reasons of begging the question.
naphra2 4 years ago
Hi naphra2, because we can only shame something that knows that it can be wrong. the computer can say to itself "gee, I really shouldn't let that program crash, that's against the rules"....... there are many ways we can control somebody--pay them money, beat them, shame them. These are distinct ways of control; we must draw the distinction in precise speech.
randyhelzerman 4 years ago
Yes I see that. Let me digest. I agree, certainly, that shaming is just one particular way of controlling.
naphra2 4 years ago
(wrt complexity) Yes, I can agree also with you Randy, there's probably a qualitative difference involved as well, not any measure of complexity will do here.
naphra2 4 years ago
"Do you think that the ambien driven sleep driver drove by choice?"
Yes. _And_, i think that the sleeping driver is, in an important sense, not the same person as the waking driver. And a consequence of that is that punishing the waking driver for the sleeping driver's crimes would be unjust.
bitbutter 4 years ago
Yup, quite so, seconded.
naphra2 4 years ago
Hi naphra2, I don't think there are two entities; I think there is one entity which either has some modules turned on or some modules turned off. I'm trying to analyze the function of the modules which are turned off in the ambien drivers case. What is it about the difference which is so special???
randyhelzerman 4 years ago
Yes, and I'm willing to go along with that to an extent (also some of it is a matter of semantics), but I'm slightly uneasy with an idea of a "free will -bearing module". But I do see what you're getting at (I think).
naphra2 4 years ago
Hi bitbutter, if there were two different entities, its merely a case of one being the subset of the other. they use the same neural circuts to, say drive, eat, open fridge doors, etc and they occupy the same spatio-temporal space. Its not two different entities, its one entity with something turned off. I'm interested in what has been turned off. That is what I'm focusing on.
randyhelzerman 4 years ago
"Its not two different entities, its one entity with something turned off. I'm interested in what has been turned off. That is what I'm focusing on."
I want to distinguish between entity (system?) and person. The driver and the waking ambien user are the same entity-with different modules active, but I consider the two different people (this way lies a foggy grey area: at what point does an amnesiac war criminal become a different person to the one who committed the crime? etc).
bitbutter 4 years ago
I do a bit of bird watching, and to attribute human like behaviors to theirs is like a siren song to the soul, but I suspect far from the truth.
grenangle 4 years ago
Randy, You are fading away on me :)
KelsoFabulous 4 years ago
Hi Fab; looks like nobody can make any sense of what I'm saying; so you're in good company ! :-)
randyhelzerman 4 years ago
I think bitbutter's point _is_ anthromorphism and that it is intentionnal: what he wants to show is that a definition of choice would apply to hypercomplex computers and since computers exist in a whole scale of power and memory sizes and that since cannot really draw an arbitrary line anywhere, you find yourself considering anything as choice making.
maksiiiskam2 4 years ago
maksiiiskam2: exactly so. Anthropomorphism is only a crime if the attribution of human characteristics to non-human things is inappropriate. It's yet to be shown that attributing choice making power to complex machines is inappropriate (and I think that's a hard case to make because people already use the word 'choice' in that way).
bitbutter 4 years ago
Hi bitbutter, suppose I hit somebody so hard that they die. Now suppose a dot matrix printer's spring pops out of the printer and hits somebody so hard that they die. I get charged with 1st degree murder. Does the dot matrix printer? Of course not. _that_ is the difference I'm trying to emphasize here.
randyhelzerman 4 years ago
Sure there's a big difference. But lets soup-up the dot matrix printer! Give it sensors, a memory, and a means of integrating the two to create new behaviour; If the dot matrix's was fitted with a 'brain' of such complexity that for practical purposes the machine's output could not be accurately predicted for any given input, we might have a machine that it would make sense to punish.
bitbutter 4 years ago
YES!
2bsirius 4 years ago
The Ambien sleepdriver case is a fascinating one, nevertheless I feel inclined to ask in the same vein, would you hold a person sleeping in the back seat culpable for a driver's offenses? i.e. who is to say there's only one distinct responsible choice-maker within one body?
naphra2 4 years ago
Hi naphra2, sure, I'd put the asleep person in the "back seat" as it were. Your question is there more than one distinct responsible choice-maker within one body, which is a good question--basically, is there somebody else who is in the driver's seat here? There is some entity which is doing the driving, and that entity is distinct from the person which thee body belongs to--but is that entity a responsible, rational, agent? (cont)
randyhelzerman 4 years ago
(cont, to naphra2) here's how I would approach it: in order to expalin the behavior of the entity, do I have to describe the cases of its behavior in normative terms? For the ambien sleep driver, I don't think so; I think we can explain the behavior in purely descriptive terms, as the satisfaction of an urge. Compare it with, say, an amoeba, which hunts for food in a drop of water. We wouldn't say its reasoning--what does it reason with? Same goes for the ambien-entity.
randyhelzerman 4 years ago
Hm, I'm not sure if it's so obvious. Ok, perhaps the sleepdriver can be described as satisfying an urge (but then again, what urge does shifting gears satisfy? Rather, it's almost normative, as it's "good for the engine" to shift gears. Anyways.) But consider something like somniloquy where people at times narrate their dreams to vivid detail (Dion McGregor, say) or express opinions or emotions. It's not clear to me it could explained descriptively.
naphra2 4 years ago
Or rather, maybe could be, but not to any better extent than explaining normal everyday waking-hour choices descriptively.
I guess I'm partially getting at what Az pointed out in his response, that the notion of will seems to me still left a little bit in the haze. Do we have A will, or a society of competing wills in our head, a la Minsky?
naphra2 4 years ago
(Clips of Dion McGregor's somniloques seem to be available on Amazon if you want to have a peek, search Amazon for "Dion McGregor Dreams Again". It's really hilarious stuff, and I'm unable to conjure up any kind of straightforward descriptive explanation for some of it. ("No of course you can't play the radio!! Don't play that radio!! You're waking Mom!!"))
naphra2 4 years ago
I think you're confusing athropomorphism with pareidolia.
CousinoMacul 4 years ago
Hi Javier, good point :-)
randyhelzerman 4 years ago
We are comfortable to say that complex machines make real choices. My contention is that humans belong to this group of complex machines. It's not clear what pareidolia has to do with this.
bitbutter 4 years ago
bitbutter,
The examples of anthropomorphism that Randy gave in his video (seeing the ace on Mars, etc.) were actually examples of pareidolia.
CousinoMacul 4 years ago
ahh i see. srry!
bitbutter 4 years ago