Newcomb's Problem Revisited

Philosophy Forums Philosophy
1

There’s a new youtube video by Veritasium that’s introducing loads of people to this problem in decision theory:

This topic has been discussed before on the forum, but perhaps it’s time for a new discussion.

You face two boxes:

Box A is transparent and contains £1,000.

Box B is opaque and contains either £1,000,000 or nothing.

You must choose either:

  • Take only Box B, or
  • Take both boxes A and B.

You are told that before you arrived, a highly reliable predictor examined the situation and predicted what you would choose. You know and accept that he’s highly reliable, near perfect even.

  • If the predictor predicted that you would take only Box B, it placed £1,000,000 in Box B.
  • If the predictor predicted that you would take both boxes, it left Box B empty.

The prediction has already been made and the boxes have already been filled (or not filled).
The predictor’s accuracy is extremely high—correct in almost every previous case.

So do you take one box or two?

  • I take One Box
  • I take Both Boxes
0 voters
2

I like boxes, I take the two, thanks

3

.
I went for one box.. because the £1,000,000 would be near-guaranteed to be contained within it.

4

Box B.

£1,000 (similar to 1,000 US dollars) is not transformative.

Taking only box B assumes a large chance of perhaps 99% depending on the reliability of the predictor (possibly Ai.) That’s 99% odds of getting a transformative, substantial sum of money (£1,000,000).

In the case of taking both box A and box B, the odds are vague and undefined, for all we know there is only a 0.1% chance of the £1,000,000 being within box B in that case.