User Controls
Determinism
-
2018-01-24 at 2:12 PM UTCwouldn't you prefer it stay in one thread rather than ejaculate out all over the board?
-
2018-01-24 at 2:36 PM UTC
Originally posted by aldra Going back on what I said in the last post, maybe random counts as 'free will'? As in will that isn't constrained, even if it isn't coherent.
The semantics aren't all that important to me though; I just don't see any alternative to either a deterministic or a random universe, and by extension the behaviours of anything that exists within it. Any argument for free will if you accept the above just seems like a way to try to preserve a sense of 'self' where there doesn't really appear to be one.
It's not to preserve a sense of "self", that is a symptom of creating a meaningful framework for attribution and responsibility. If you have a waffle iron that makes particularly good waffles, you wouldn't say "wow the big bang makes a damn good waffle". In much the same way, we can have an argument about the nature of agency, but first you must acknowledge the logical validity of the framework for free will that I'm offering (I.e. an agent's will is free when it's unconstrained by any outside agent).
One more thing to note is that my framework can actually be applied both to my compatibilist conception of free will, which works in a universe that has causal rules, but it can also describe the vague "general" notion of free will. That's why the supposedly "general notion" of free will is irrelevant; we can move forward with my framework in either case!In the case of your robot, it still follows the same pattern as a person - sets of behaviours that are created deterministically (either by design or by the results of previous behaviours/decisions), which are then affected by external (random) stimuli. You could (and probably will) argue that the quantum matter it derives its random seed from is internal, but it doesn't change the fact that it's random and therefore results in a randomised behaviour
Then you're playing a weird game of semantic inconsistency: having a new deterministic, causal chain beginning from a random seed inside the robot's brain (let's say that it only rolls the quantum dice one set of times to determine the robot's "personality" and how it reacts to the environment for the rest of it's life) is no more "random" than the causal chain of the rest of the universe. You could probably look at the robot's personality and construction and, with enough information about how it behaves and what stimuli it will receive, predict exactly how it will act until the end of the universe. But that causal chain would have begun separate from the causal chain of the rest of the universe. -
2018-01-24 at 5:04 PM UTC
Originally posted by Captain Falcon It's not to preserve a sense of "self", that is a symptom of creating a meaningful framework for attribution and responsibility. If you have a waffle iron that makes particularly good waffles, you wouldn't say "wow the big bang makes a damn good waffle".
You wouldn't say it because there's nothing to be gained from the extra information - you could literally preface anything with "a long, predictable chain of events did this", and for that reason it's not necessary. Also, most people don't think about determinism vs. free will when they're making waffles.In much the same way, we can have an argument about the nature of agency, but first you must acknowledge the logical validity of the framework for free will that I'm offering (I.e. an agent's will is free when it's unconstrained by any outside agent).
let me put it this way. AI is more or less just a huge series of conditional statements that can be overwritten at runtime, ie. the AI encounters external stimuli and updates itself.
If the programmer is able to control the stimuli that the AI experiences, and as a result is able to predict exactly how the AI will behave, indefinitely, does it have 'free will'?Then you're playing a weird game of semantic inconsistency: having a new deterministic, causal chain beginning from a random seed inside the robot's brain (let's say that it only rolls the quantum dice one set of times to determine the robot's "personality" and how it reacts to the environment for the rest of it's life) is no more "random" than the causal chain of the rest of the universe. You could probably look at the robot's personality and construction and, with enough information about how it behaves and what stimuli it will receive, predict exactly how it will act until the end of the universe. But that causal chain would have begun separate from the causal chain of the rest of the universe.
That causal chain is still part of the universe. It doesn't magically start when the robot boots up - single celled organisms evolve into fish. Mammals emerge. Humans kill everything. Humans dominate and eventually build robots.
I don't see that it makes a difference regardless though - random behaviours emerge from random behaviours. Birth, creation, death, destruction, rebirth - there's nothing to differentiate one state from another. They're all built from the same components, and they're all just a prelude to something else -
2018-01-24 at 5:06 PM UTC*honestly having trouble trying to find a way to phrase why I think even the concept of free will doesn't make sense
-
2018-01-24 at 5:14 PM UTC
-
2018-01-24 at 5:14 PM UTC
-
2018-01-24 at 6:13 PM UTC
Originally posted by aldra You wouldn't say it because there's nothing to be gained from the extra information - you could literally preface anything with "a long, predictable chain of events did this", and for that reason it's not necessary. Also, most people don't think about determinism vs. free will when they're making waffles.
I'm asking you to think about the conceptual weight of the idea: you consolidate that chain of events into the entity of "waffle iron", even though that doesn't really mean anything on any physical level.let me put it this way. AI is more or less just a huge series of conditional statements that can be overwritten at runtime, ie. the AI encounters external stimuli and updates itself.
I understand that. But you couldn't predict how the AI's "weighting" will turn out no matter who you are if you're generating it based on a truly random seed, like with a QRNG. That particular AI's future will be determined from events that have no causal link outside of the system we call the robot. How would you say that that system does not determine it's future from inside itself?If the programmer is able to control the stimuli that the AI experiences, and as a result is able to predict exactly how the AI will behave, indefinitely, does it have 'free will'?
No, because that would be a constraint or compulsion by another agent.That causal chain is still part of the universe. It doesn't magically start when the robot boots up - single celled organisms evolve into fish. Mammals emerge. Humans kill everything. Humans dominate and eventually build robots.
Everything is part of "the universe", by definition. But it's not caused or determined by any prior event. No amount of information could predict what the outcome of the QRNG would be. It's no more or less random or determined than "the universe" itself. How is this a point of contention? Does an entity literally have to spontaneously generate from the quantum vacuum for you to consider it free?I don't see that it makes a difference regardless though - random behaviours emerge from random behaviours. Birth, creation, death, destruction, rebirth - there's nothing to differentiate one state from another. They're all built from the same components, and they're all just a prelude to something else
I don't see how you can differentiate one from the other. This is a form of logical fallacy called special pleading. -
2018-01-24 at 6:28 PM UTC
Originally posted by Captain Falcon I'm asking you to think about the conceptual weight of the idea: you consolidate that chain of events into the entity of "waffle iron", even though that doesn't really mean anything on any physical level.
I understand that. But you couldn't predict how the AI's "weighting" will turn out no matter who you are if you're generating it based on a truly random seed, like with a QRNG. That particular AI's future will be determined from events that have no causal link outside of the system we call the robot. How would you say that that system does not determine it's future from inside itself?
No, because that would be a constraint or compulsion by another agent.
Everything is part of "the universe", by definition. But it's not caused or determined by any prior event. No amount of information could predict what the outcome of the QRNG would be. It's no more or less random or determined than "the universe" itself. How is this a point of contention? Does an entity literally have to spontaneously generate from the quantum vacuum for you to consider it free?
I don't see how you can differentiate one from the other. This is a form of logical fallacy called special pleading.
^ gay. -
2018-01-24 at 9:50 PM UTC
-
2018-01-24 at 10:05 PM UTC
Originally posted by aldra wouldn't you prefer it stay in one thread rather than ejaculate out all over the board?
Seems like there are a lot of threads talking about the same thing lately and a lot of other threads eventually turn to it
Regardless, I was pretty fucked up when I posted that, it's not that big of a deal
