Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Justified Expectation of Pleasant Surprises

10 Eliezer_Yudkowsky 15 January 2009 07:26AM

I recently tried playing a computer game that made a major fun-theoretic error.  (At least I strongly suspect it's an error, though they are game designers and I am not.)

The game showed me—right from the start of play—what abilities I could purchase as I increased in level.  Worse, there were many different choices; still worse, you had to pay a cost in fungible points to acquire them, making you feel like you were losing a resource...  But today, I'd just like to focus on the problem of telling me, right at the start of the game, about all the nice things that might happen to me later.

I can't think of a good experimental result that backs this up; but I'd expect that a pleasant surprise would have a greater hedonic impact, than being told about the same gift in advance.  Sure, the moment you were first told about the gift would be good news, a moment of pleasure in the moment of being told.  But you wouldn't have the gift in hand at that moment, which limits the pleasure.  And then you have to wait.  And then when you finally get the gift—it's pleasant to go from not having it to having it, if you didn't wait too long; but a surprise would have a larger momentary impact, I would think.

This particular game had a status screen that showed all my future class abilities at the start of the game—inactive and dark but with full information still displayed.  From a hedonic standpoint this seems like miserable fun theory.  All the "good news" is lumped into a gigantic package; the items of news would have much greater impact if encountered separately.  And then I have to wait a long time to actually acquire the abilities, so I get an extended period of comparing my current weak game-self to all the wonderful abilities I could have but don't.

Imagine living in two possible worlds.  Both worlds are otherwise rich in challenge, novelty, and other aspects of Fun.  In both worlds, you get smarter with age and acquire more abilities over time, so that your life is always getting better.

But in one world, the abilities that come with seniority are openly discussed, hence widely known; you know what you have to look forward to.

In the other world, anyone older than you will refuse to talk about certain aspects of growing up; you'll just have to wait and find out.

continue reading »

Continuous Improvement

16 Eliezer_Yudkowsky 11 January 2009 02:09AM

When is it adaptive for an organism to be satisfied with what it has?  When does an organism have enough children and enough food?  The answer to the second question, at least, is obviously "never" from an evolutionary standpoint.  The first proposition might be true if the reproductive risks of all available options exceed their reproductive benefits.  In general, though, it is a rare organism in a rare environment whose reproductively optimal strategy is to rest with a smile on its face, feeling happy.

To a first approximation, we might say something like "The evolutionary purpose of emotion is to direct the cognitive processing of the organism toward achievable, reproductively relevant goals".  Achievable goals are usually located in the Future, since you can't affect the Past.  Memory is a useful trick, but learning the lesson of a success or failure isn't the same goal as the original event—and usually the emotions associated with the memory are less intense than those of the original event.

Then the way organisms and brains are built right now, "true happiness" might be a chimera, a carrot dangled in front of us to make us take the next step, and then yanked out of our reach as soon as we achieve our goals.

This hypothesis is known as the hedonic treadmill.

The famous pilot studies in this domain demonstrated e.g. that past lottery winners' stated subjective well-being was not significantly greater than that of an average person, after a few years or even months.  Conversely, accident victims with severed spinal cords were not as happy as before the accident after six months—around 0.75 sd less than control groups—but they'd still adjusted much more than they had expected to adjust.

This being the transhumanist form of Fun Theory, you might perhaps say:  "Let's get rid of this effect.  Just delete the treadmill, at least for positive events."

continue reading »

Harmful Options

23 Eliezer_Yudkowsky 25 December 2008 02:26AM

Previously in seriesLiving By Your Own Strength

Barry Schwartz's The Paradox of Choice—which I haven't read, though I've read some of the research behind it—talks about how offering people more choices can make them less happy.

A simple intuition says this shouldn't ought to happen to rational agents:  If your current choice is X, and you're offered an alternative Y that's worse than X, and you know it, you can always just go on doing X.  So a rational agent shouldn't do worse by having more options.  The more available actions you have, the more powerful you become—that's how it should ought to work.

For example, if an ideal rational agent is initially forced to take only box B in Newcomb's Problem, and is then offered the additional choice of taking both boxes A and B, the rational agent shouldn't regret having more options.  Such regret indicates that you're "fighting your own ritual of cognition" which helplessly selects the worse choice once it's offered you.

But this intuition only governs extremely idealized rationalists, or rationalists in extremely idealized situations.  Bounded rationalists can easily do worse with strictly more options, because they burn computing operations to evaluate them.  You could write an invincible chess program in one line of Python if its only legal move were the winning one.

Of course Schwartz and co. are not talking about anything so pure and innocent as the computing cost of having more choices.

If you're dealing, not with an ideal rationalist, not with a bounded rationalist, but with a human being—

Say, would you like to finish reading this post, or watch this surprising video instead?

continue reading »