actually after a cold one it kinda make sense. you describe determinism.
with a small twist which im not sure yet if its a twist at all.
Does True Randomness Actually Exist? ( ^&*#^%$&#% )
Sort:
Oct 14, 2021
0
#2703
Optimissed wrote:
Elroch wrote:
Those who prefer non-optimal choices haven't noticed that that makes the choices optimal according to their chosen criteria.
For example a person who plays chess and throws in deliberate inaccuracies (none of that sandbagging here of course!) in order to make their opponent happy is choosing optimal moves based on a criterion that values the opponent's happiness.
That's a question on which I pondered as a child. I thought about it for several days when I was about nine. You're incorrect because in order to reach such a conclusion, you're actually defining "optimal" as "that which is chosen".
Not quite. You should read what I said. There is of course no universal definition of optimal, but what I said was "optimal according to your criteria". Your criteria are the gamut of conscious and unconscious conditions that determine the action in question.
To put it another way. "that which is preferred" = "that which is chosen".
It's tautologous, so not that exciting. You were right to say it is a definition, but the term is "optimal according to your criteria", not "optimal".
It's possible to use reductio ad absurdum to refute it.
Really? I think not, but have a go!
However, it's possible to make the counter-argument that the reductio ad absurdum is optimal
?
because it's being used to refute the argument. But that doesn't hold for all cases. It only holds for the example. Life extends beyond proofs of things and there's no possible reason to assume that all our actions are optimal because we do them. It's an argument that feeds only on itself.
This argument has bitten off a piece of itself.
Oct 14, 2021
0
#2704
Optimissed wrote:
Elroch wrote:
Isn't that more what could be called "purposeful behaviour"? Where you somehow assess choices based on some sort of evaluation of perceived consequences and use this evaluation to make the choice.
A very mundane example of the same can be found in many technologies. For example, a robot that detects which its battery is low, seeks and finds a power socket and recharges itself.
It is psychologically attractive for us to focus on the "freeness", but wouldn't it be accurate to say that what matters is we make the _best_ choices we can, based on our our criteria?>>
In common with other animals, we only need to make the best choices we can make when it's important to do so. Someone who always makes the best choices in any circumstance is "less free" than someone who doesn't. Their behaviour might be considered to be "determined" to a higher degree than that of others, whose behaviour is less predictable but not necessarily random. It might be for this reason that many scientists believe that there's no free will; human behaviour being determined.
In a sense this removes the freedom, except in the case where the choice doesn't matter. Even there while we may like to have the option to choose between equally beneficial choices, by definition it would be no disadvantage to be forced into some one of those equally best choices.
It might be an advantage to choose sub-par options, some of the time.
There is a truth hidden behind the imprecise wording here. If an option is "sub-par" on any occasion, it is obviously inferior by definition to some other option so NOT advantageous.
But the truth I am reminded of is the notion of randomised strategies such as Nash equilibria. Here it is not that an option is "sub par", but rather that if your strategy is defined by a probability distribution of actions, the strategy that does best against the best counter strategy does not always choose the same option in the same situation (because such a strategy can be exploited).
A nice example is the serve in tennis. Always doing the same serve is clearly exploitable. Randomising the placement somehow is better. The optimal randomisation depends on the relative difficulty of dealing with each of the options.
In this context, it is incorrect to thing of one of the serves as "sub-par". It is the distribution of serves that is the thing which has optimality.
This complication only arises when there is a conflict between purposeful agents.
Viewed like this it seems the ability to make good choices matters a lot more than "freedom" (especially when the freedom enables the possibility to make what by our own criterion is a worse choice).
To some people.
elroch, say you have two glasses of water in front of you, and youre thirsty. so to say that youre facing an equal choice. thats where your random choices kicks in? did i understand you all right?
in las vegas energy (money) is transferred (minutely affecting the universe and moreso the earths) w/ the roll of dice and other games a CHANCE (can u say random ?...its easy. puff up ur lips then pull ur tongue offa ur mouth roof then hum w/ ur mouth closed).
is there a better way show a portion of existence is random ?
Oct 14, 2021
0
#2707
Bercher wrote:
elroch, say you have two glasses of water in front of you, and youre thirsty. so to say that youre facing an equal choice. thats where your random choices kicks in? did i understand you all right?
That is an example of the class of cases where randomness does not conflict with optimality. This example is uncomplicated by some opponent trying to second guess you!
I am trying to clarify in my mind what it is that in some situations makes randomised choices better, more than merely optional (like in the case of the glasses of water).
The key is some sort of feedback between your choices and the environment, so that you need to balance the simple goodness of choices against the negative feedback that is the result of being predictable. The notion of "predictability" involves some sort of model of your behaviour existing, and the environment changing its behaviour in response. Of course all the examples of this involve purposeful agents.
Oct 14, 2021
0
#2708
I believe Kurt Godel solved this question about eighty years ago , it was something like 50/50 or maybe 60/40. Who can remember so long ago. The o in Godel should have an umlaut but I couldn’t find one. He gets mentioned in GEB by Douglas Hofstadter.
Thee_Ghostess_Lola wrote:
in las vegas energy (money) is transferred (minutely affecting the universe and moreso the earths) w/ the roll of dice and other games a CHANCE (can u say random ?...its easy. puff up ur lips then pull ur tongue offa ur mouth roof then hum w/ ur mouth closed).
is there a better way show a portion of existence is random ?
to us a lot of our existence is random. thats just stating the obvious. same as a roulette is random to the player. but when it comes to FW what we really want to know is if our choices actually makes a different, or are we just actors in the show of the universe.
if we do have free will, than our choices really make a difference. and if we dont.. then whatever.
"I am trying to clarify in my mind what it is that in some situations makes randomised choices"
yes. will be nice if you come up with your own example. all you have to demonstrate is a real life situation with your choice being random and independent from any criteria.
Oct 15, 2021
0
#2711
You appear to have misunderstood. I explained I was just considered exactly what it is that means no deterministic strategy is optimal (so only randomised strategies are optimal).
The answer is that all deterministic strategies can be sub-optimal where there is some sort of negative feedback that penalises predictable behaviour (by using the information about that behaviour in choices).
This typically occurs when there are purposeful agents that are in some sense competing (I can't think of a single other example - can anyone else?)
Oct 15, 2021
0
#2712
The genome has the ability to learn too. It is a slow process, stochastically exploring a massive space of variation driven by the filter of natural selection through empirical feedback called fitness.
Optimissed wrote:
Bercher wrote:
it doesnt make much sense to me either, but im curious where he's going with that.
the "standard argument" is from a reductionist POV.
I would think nowhere. There's no possible proof of determinism and Elroch himself doesn't accept determinism so I think he was only trying (and failing) to win a "clever" point.
maybe you nailed it right there, or maybe he can make sense out of it. im not sure yet.
elroch, you believe that we are not free to make choices, and our choices are always govern by some given criteria. is that right? and if so.. can you give a real life example of a choice you ever made that is not deterministic?
I think he is still trying to work out the hundredth monkey phenomenon.
absolutely not! I believe in genuine free will, and none of the apologetic definitions of it.
but i also believe in objectivity, and try to draw a very fine line between my personal beliefs and accepting other beliefs. thats why i strive to understand others.
stating that elroch is a clever guy is just stating the obvious.
if it seem that way, then i must have done something wrong. i was arguing against "science proved determinism to be false" thats just complete rubbish and you know it.
so whos burden izzit to prove ? for either chessers or free willys ?
Oct 24, 2021
0
#2719
Causality means that a free choice can only affect what happens in its future light cone (or to slightly over-simplify, in its future).
It is important to realise that this does not just apply to definite consequences. Causation is about affecting the statistics of what happens.
I use the phrase "free choice" here because it is not true that there cannot be information that informs you about things that are not in its future. For example, if there is a box with a penny in it, but it could be either way up, and you look at this penny at some point in time, the observation informs you about the state of the penny (secretly filmed say) at some time in the past. The future information does not cause the past information, even though there is the exact same statistical relationship as causation (heads now implies heads in the past, tails now implies tails in the past - just the same as with the times swapped).
An example of a choice would be whether we cook the penny or freeze it, a choice that occurs at the future time. This choice provides no information about whether the coin was heads or tails in the past.
EDIT: the above is really just a preamble - everything that matters is below!
It is more relevant to think of the information flows in and out of a closed system that contains the box containing the coin. The coin state is not new information to the system (containing the box), it is information that is within the system. But a choice to cook or freeze the coin can be due to new information entering the system to decide what to do. It is because it is new information, independent of the system, that it cannot inform us about the past state of the coin.
If this was all formalised, it seems it would come down to this notion of independent subsystems containing information and flows of information into those subsystems. Causality says information entering a closed system can only have a statistical effect on events in the future light cone of where the information enters the system.
In the real world, there would be practical difficulties in ensuring information came from outside a system: it seems to me that the best that can be done is to achieve this for practical purposes rather than strictly (due to the pedantic necessity to track information and closure through all cosmological history). That being said, an excellent source of external information would be observations of the cosmological microwave background (least significant bit of digital-analog conversion is a nice choice), which is information that should be independent for more than 13 billion years into the past and we can be essentially 100% confident is genuinely independent of the system it is entering.
[An important example is the choice of observation in a Bell's experiment. This choice cannot affect the statistics of the observations at the other end of the experiment, because that is not in the future light cone of the choice. But the observation is statistically related to the observation at the other end, in a subtle way revealing entanglement in the overall statistics of the results of many different combinations of (causally independent) observations].
it doesnt make much sense to me either, but im curious where he's going with that.
the "standard argument" is from a reductionist POV.