summaryrefslogtreecommitdiffstats
path: root/tmp/Newcomb paradox
blob: c7da22b5524962080c4fb2afa2846a2b2761951b (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
==================
Newcomb's Dialemma
==================

Newcomb's paradox was thought up by a researcher named Newcomb; it was first
explored and written up by Robert Nozick in the 1969 paper 
"Newcomb's Problem and Two principles of Choice". 

The Situation
-------------
As narrated by an all knowing "predictor"::

    I am going to give you a choice. It is important to know that I really
    pretty much know what you are going to do. I have been watching their whole
    life and am additionally an immortal being; i've been doing this a long
    time and always guess correctly. It's also important to know that I am
    unbiased and don't care which decision you make, I have nothing to gain
    either way.

    Here are two boxes: a large and a small. The small has a 10 shekel coin
    in it (show everybody). The large one may or may not have a thousand
    shekels in it; you don't know. Your choice is to either take only the
    large box or to take both the large and small boxes. The twist is that
    I already knew which decision you will make and decided whether or not
    to put the $1000 in the large box or not based on that knowledge.
    If I knew you would "two box", then I left the large box empty. If I knew
    you would "one box" then I filled it. 

Dominance Mindset
-----------------
Regardless of what decision was made previously, and whether or not there
is anything in the large box, the person is better off taking both boxes;
either they will get just $10 (better than none) or $1010 (better
than $1000). So two-box.

Trusting Mindset
----------------
The predictor is pretty much always right so we can just ignore the 
possibility that they are wrong. In this case, choosing to one-box
implies that the Predictor knew you would and you get $1000; 
choosing to two-box implies that the predictor knew you would and you
only get $10.

The predictor doesn't even have to be perfectly accurate; say they are
90%:
If you one-box, your expected value is $900.
If you two-box, your expected value is $110.

Discussion
----------
It's disputed whether this is a paradox, and there are many deeper arguments
that I don't have time to go into here. Ultimately, I am a one-boxer 
though this is something of a minority position.

Afterword
---------
The person who taught me this paradox, Professor Augustin Rayo, a
two-boxer, then had this to add. He was talking with his one-boxing friend
and accused her of letting irrationality undermine her logic: she is so
optimistic that if a statement S is unprovable, but it would be nicer if S
was true than false, then she pretens that S is proven. So basically, even
though there is no rationalization, she will accept a statement "just
because it would be nice", and this isn't how logic works. To which she
replied "but wouldn't it be nice if it was?".