2 kids found 2 similar boxes. They opened box 1. It contained a cookie. Kid 1 immediately grabbed it and started eating it, saying “the other box probably has a cookie too; that one is yours” (proposition A)
Conversation
By the time kid 2 got box 2 open and discovered it contained a lump of coal, kid 1 had eaten the cookie.
Between the first bite of the cookie and the discovery of the lump of coal, both kids inhabited a reality where proposition A was assumed to be true.
1
4
This is what it means for a reality to be constructed out of the benefit of doubt. The brief period where Prop A *might* have been true was also the period when kid 1 was *really* eating the cookie.
1
6
In modal logic terms, the cookie was eaten in all possible worlds (P: ▢ the cookie is eaten by kid A) but there was a second cookie only in *some* possible worlds (Q: ⋄ There is a cookie). In between was world W where both P and Q were true for period T.
2
1
Replying to
In your model, in W Q was ASSUMED AS PROBABLY TRUE, it was never true; kid 2 gambled on what the other box contained
1
Replying to
Probability theory is a an unnecessary level of epistemology that is neither necessary not sufficient to account for modal structure of subjective reality. I’m trying to describe at a level that works even for belief dynamics of babies and chimps.
1
Replying to
I think de finetti answers this perfectly, and i agree with him that even simple living beings operate on subjective probability estimates all the time
1
Replying to
But not consciously
Just as modulation of emotional states like caring, excitement etc via dopamine regulation mainly. “Probabilistic belief” maps to things like “engaged” or “apathetic” and is not determinative (other factors modulate those states too)
2
Replying to
There is no need for coscience to enter the picture, i agree. Living beings are wired to gamble
1
Replying to
They’re wired to behave in ways we tend to narrativize as “gambling”
It’s like pretending cats and dogs can understand what we say in human language because they respond appropriately sometimes
Replying to
We model it as gambling because it works as a descriptive and predictive model of behaviour. Risk taking under uncertainty is basically every day in life
1
Replying to
Sure. You’re modeling for predictive success. I’m modeling for descriptive accuracy of subjective experience. “What is it _like_ to hold what logicians call a probabilistic belief”
1
1
Show replies

