An episode of the Noddy animated series has the following plot.

Noddy needs to go pick up Martha Monkey at the station. But it's such a nice, sunny day that he would prefer to play around outside. He gets an idea to solve this dilemma. He casts a duplication spell on himself and his car and tells the duplicate to go fetch Martha while he goes out to play. Later, Noddy is out having fun when he suddenly spots his duplicate. It turns out that the duplicate also preferred playing outside to doing the errand so he also cast a duplication spell. Then they see another duplicate, and another...

I think this story makes for a nice simple illustration of one of our perennial decision theoretic issues: When making decisions you should take into account that agents identical to yourself will make the same decision in the same situation. A common real-life example of the Noddy problem is when we try to pawn off our dietary problems to our future selves.

New Comment
26 comments, sorted by Click to highlight new comments since: Today at 1:48 PM
[-]Thomas12y400

JERRY (SEINFELD): I never get enough sleep. I stay up late at night, cause I'm Night Guy. Night Guy wants to stay up late. 'What about getting up after five hours sleep?', oh that's Morning Guy's problem. That's not my problem, I'm Night Guy. I stay up as late as I want. So you get up in the morning, you're ..... (?), you're exhausted, groggy, oooh I hate that Night Guy! See, Night Guy always screws Morning Guy. There's nothing Morning Guy can do. The only Morning Guy can do is try and oversleep often enough so that Day Guy looses his job and Night Guy has no money to go out anymore.

You can punish and intimidate your future self, as you can see.

What about your past self? If Night Guy can predict what Morning Guy will do, Morning Guy is effectively threatening his past self.

So awfully true.

With school just getting back in, I find myself continuously saying "I'll do my work at X o'clock" where X is always multiple hours in the future, and then when I get to that time, I find myself saying the same thing. It's like I forget that my future self is just like me.

So that this comment has more than epsilon information, I'll also provide the heuristic I'm going to use this semester. Hopefully it works, etc., etc.:
Every time I find myself pushing the "get to work time" on my future self, assume I will not work at all that day. Maybe this motivates me to work immediately, maybe not. If not, at least I don't get the guilt and stress of deluding myself into thinking I'll get to work X hours from now.

this is literally the best solution I could think of

[-]Thomas12y120

Your future self will have less time to do the tasks. He will at least not be able to postpone those with the same rationalization as you can.

And this your present self knows.

It seems to be a problem of "desired processes" vs "goals". Picking Martha Monkey up from the station is a goal, while playing outside is a desired process. If one Noddy picks up Martha Monkey from the station, all of the Noddys' desires for Martha to be back from the station are fulfilled. However, if a Noddy plays outside, this only fulfills that Noddy's desire to play outside.

This looks a lot like the "tragedy of the commons" idea so often referred to in economics, when framed in the sense of multiple simultaneous instances of the same person. This is the first time I have really understood how TDT works.

There was also a Calvin and Hobbes storyline in which Calvin, not wanting to do his homework, makes a time machine and travels a few hours into the future so he can collect his already-done homework and take it back with him. Of course, when he arrives in to the future, the homework still isn't done, because he spent the time between then and now goofing off...

C&H also had the 'duplicator' storyline (about cleaning rooms and going to school, IIRC), which was the OP but much more so.

I read a lot of C&H growing up, and looking back at it, I'm surprised at how many interesting ideas it contains. I wonder how much of my present self was shaped by having these ideas implanted at age 8 or 9...

There was also a Calvin and Hobbes storyline in which Calvin, not wanting to do his homework, makes a time machine and travels a few hours into the future so he can collect his already-done homework and take it back with him. Of course, when he arrives in to the future, the homework still isn't done, because he spent the time between then and now goofing off...

I would think the problem there is that there isn't any instance of Calvin in the time stream at that point - goofing off or otherwise.

The correct decision there is to do the homework then take it back in time. Which, if you (would) decide to actually do you will never have to do. You just get the homework delivered to you, be glad you are the kind of person who can make good decisions and then go ahead and pass on your good fortune to your past self. This latter is itself a Newcomblike choice making the decision problem here twofold. One decision you never actually make while the latter decision you make despite it having no future benefit.

What happens next is that Calvin and his future self decide to go back in time halfway and force THAT Calvin to do it. Of course, the Calvin of that time doesn't want to do either; the three of them all end up fighting each other, while Hobbes ends up writing the story that Calvin was supposed to do for homework. Calvin then turns it in without reading it... he gets an A, but he's mad at Hobbes anyway because, when he gets it back, Calvin realizes that the story (which was about Calvin trying to use his time machine to get out of writing a story for homework) made him look like an idiot.

Best line in the story (during the fight among Calvins): go ahead and hit me, it'll be the future Calvin who hurts!

Nice illustration. It also lends itself to drawing an analogy between a solution to the Noddy Problem and real life akrasia. In the Noddy problem, the obvious solution is to make sure that the duplicate is not identical (or rather, that their environment is not close enough to yours in the relevant ways). One way to do this is by changing incentives -- you know that you would pick up Noddy if your life depended on it, so you can create a duplicate that believes that it will explode if they don't do the job. In the real world, you can apply the same sort of solution by using self-binding to coerce your future selves into doing your bidding.

Another similar point is made in the xkcd on cryogenics.

Most people link to XKCD, but I think SMBC has tons of pictures relevant to LW topics. (Some random choices.)

Frankly, XKCD strikes me as awful political art.

Frankly, XKCD strikes me as awful political art.

Frankly your opinion seems to indicate ignorance. The vast majority of XKCDs are not political in any sense. It is nerd orientated humor which only seldom strays into humor for the benefit of nerd politics. Call it bad art if you don't appreciate the jokes or insights contained therein but just doesn't fit in the category awful political art at all.

I think MileyCyrus meant it in the sense of the linked Yudkowsky essay, in which he gives atheistic hymns as an example. In other words, people like xkcd because of the applause lights and in-group signalling, not because of the quality. For what it's worth, I disagree, though I might describe some of the lower-quality xkcd strips this way.

I think MileyCyrus meant it in the sense of the linked Yudkowsky essay, in which he gives atheistic hymns as an example.

I used the same sense. It does not apply.

Political doesn't necessarily mean "Republican vs Democrats". In XKCD's case, it's "STEM majors vs liberal arts majors."

Political doesn't necessarily mean "Republican vs Democrats".

That is true, trivially true and it surprises me that I have to repeat that yes, it is what I assumed when I initially rejected your accusation. I thought "not political in any sense" was about as explicit as I could be without just sounding awkwardly verbose.

In XKCD's case, it's "STEM majors vs liberal arts majors."

I cannot imagine how anyone who read a significant sample of XKCD comics would claim this. It is about in jokes - witty (or intended to be witty) references to ideas that are funny if you happen to either know about the technical factoid or are interested enough in that kind of thing to google it. If anything it is far too well, self absorbed in it's own STEMiness for it to bother being political much along the STEM/arts axis. It doesn't even seem to bother to try to score points along the nerds/copyright axis. It takes its cheap shots at the expense of nerds almost exclusively.

Take a look at 863, or the alt text of 764. Liberal arts majors are identified as an outgroup in XKCD's "warning" at the bottom of each page. It's as transparent as Dr Pepper 10 saying "It's not for women!!"

If you only have 2 examples out of 1002 that even debatably support your point, your overall characterization is an absurd overgeneralization at best.

How is 863 against liberal arts majors? It seems to be about lazy naive undecided students, vs people seriously interested in their major.

[-][anonymous]12y40

The warning at the bottom says, "This comic occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors)."

So yeah, hardly transparent.