Looking at intuition vs reality using philisophy or science (probably only philosophy)

[ryan]

 Bayesian Inference

You start with a prior between 0 and 1, and then update it according to the information you find. Even if the prior is really, really, really, small, if this persists for long enough, the rational conclusion will update the posterior probability to be more and more likely.

Ah, Thanks!

 

[Speakpigeon]

Do you think that evolution had a large enough sample to be correct on whether something unknown should end or not? I don't think so, which means it's wrong.

Clearly, the evolution on Earth of the human species specifically would say little about conditions somewhere else in the universe, or in the whole universe overall if you're looking for a universal law.

However, it's not '"necessarily" wrong. We just can't say.

It will be wrong if the 'topology' of the environment on Earth, relatively to the human species, has anything significant which would distinguish it from conditions elsewhere. One reason for example for a significant difference might be the scale of the features in the environment which are relevant to our intuitions about repetition/continuation of the features. It's no coincidence you picked up the shrub example. Shrub-size features have probably been much more perceptible to our species during our evolution than viruses-size features or galactic-size features. The latter two may not repeat or continue like shrubberies typically do on Earth.

And do you include behaviours in the features to be considered? Or bare properties, such as colour, density or texture?

I guess it's easy to see how your interrogation couldn't possibly have any rational answer. I'm not even sure it really makes sense at all!



**************************************************

But if it's right, then it must be saying something very profound about the unknown and only the unknown as it enters consciousness.

The unknown as it enters consciousness? What that's suppose to mean?!

Oh wait! Let's see if Subsymbolic knows the answer to that.
EB

 

[Speakpigeon]

About the questions numbered 2 then I think it is reasonable: most shrubberies have a limited size. The longer you walk, the leds is the probability that this shrubbery is similar to the shrubberies you are familiar with and this the probability that it is a very big shrubbery increases.

Yes, the use of shrubbery in the thought experiment refers us to something which previous experience indicates to us is limited in size.

The fact that we should have any experience at all as to shrubberies seems to me to constitute a bias working against a rational view as to whether our intuitions could be true of a universal law.

Another, slightly different example might be......where the persistence of something (a tossed coin repeatedly landing on a head for example) might lead us to think that the series must end at some point (and in fact in the gambler's fallacy, we can take the persistence to be actually increasing the likelihood that it will change on the very next toss).

Repetition or continuation in the case of tossing a coin may not follow the same rules as repetition or continuation in the case of shrubberies. And it's seriously worse for the general case.

I'm not even sure anyone could articulate what the general case is to begin with. It's one thing to talk of the specific and very simple case of shrubberies and coins but just imagining what the general case might be seems much more difficult, even beyond the power of mere mortals like I presume we all are here.
EB

 

[Speakpigeon]

Thinking identical coin flips will end is just as wrong as thinking they will continue.

You're thinking in abstract terms. In real life situations, there's an overall distribution in the lengths of coin tossing series, just because most tossing is done by human beings, with plenty other things to do in life beside tossing a stupid coin.

Why should the shrubs be more likely to not end the more they don't end?

I don't think there's any a priori answer to that. It's an empirical matter. For shrubs on Earth, there's a particular overall distribution of shrubbery lengths but the distribution would likely be different somewhere else in the universe.

And, presumably, all series end at some point and they will all end immediately after the last term of the series, i.e. the one coming after all the others. So, it's true to say that all series end precisely immediately after the last-to-come term, i.e. that series are more likely to end the more they continue.

Again, I think there's a problem in articulating a proper formulation of the problem.
EB