Comments on: What Evidence Ease of Imagination? http://www.overcomingbias.com/2007/09/what-evidence-e.html Overcoming Bias is economist Robin Hanson’s blog, on honesty, signaling, disagreement, forecasting, and the far future. Sat, 11 Feb 2012 23:23:58 +0000 hourly 1 http://wordpress.org/?v=3.1.3 By: Tom Breton http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416040 Tom Breton Wed, 05 Sep 2007 22:11:09 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416040 Perhaps the ease of imagination heuristic is being relegated to the trash heap too soon. It certainly has its flaws, and is a rich source of bias and false belief. On the other hand, it is a powerful mental tool for exploring ideas, especially ideas that interact with that hard-to-capture context that we call reality. It is powerful for anticipating problems and opportunities inherent in an idea which might not be superficially apparent. We shouldn't be too quick to discard such a powerful and useful tool. To use it well, it seems to me we must keep in mind these things, and perhaps others that don't occur to me at the moment: <ul> <li> Argument from not being able to imagine something is argument by failed search, with all the qualifying conditions that imposes. </li> <li> Argument from ease of imagination, whether absolute or relative ease, is subject to biases from the sources that feed our imagination, particularly the more vivid sources. For example, do we find it easier to imagine something because we saw something similar from a fictional source, such as TV? Or a source that selects what it presents, such as TV news? If so, our confidence should be no higher than our confidence in the unbiasedness of that. Usually this is a fatal limitation. If we can't identify the relevant sources, we should be especially careful. </li> </ul> Perhaps the ease of imagination heuristic is being relegated to the trash heap too soon.

It certainly has its flaws, and is a rich source of bias and false belief.
On the other hand, it is a powerful mental tool for exploring ideas, especially ideas that interact with that hard-to-capture context that we call reality. It is powerful for anticipating problems and opportunities inherent in an idea which might not be superficially apparent.

We shouldn’t be too quick to discard such a powerful and useful tool.

To use it well, it seems to me we must keep in mind these things, and perhaps others that don’t occur to me at the moment:

  • Argument from not being able to imagine something is argument by failed search, with all the qualifying conditions that imposes.
  • Argument from ease of imagination, whether absolute or relative ease, is subject to biases from the sources that feed our imagination, particularly the more vivid sources. For example, do we find it easier to imagine something because we saw something similar from a fictional source, such as TV? Or a source that selects what it presents, such as TV news? If so, our confidence should be no higher than our confidence in the unbiasedness of that. Usually this is a fatal limitation. If we can’t identify the relevant sources, we should be especially careful.
]]> By: Roy Haddad http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416039 Roy Haddad Mon, 03 Sep 2007 17:06:06 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416039 It's a bias as well, because this heuristic is what many people use in place of better processes for judging likelihood (especially while in the grip of fear). Aren't many biases just over-applied heuristics? It’s a bias as well, because this heuristic is what many people use in place of better processes for judging likelihood (especially while in the grip of fear). Aren’t many biases just over-applied heuristics?

]]> By: Hopefully Anonymous http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416038 Hopefully Anonymous Sun, 02 Sep 2007 19:36:49 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416038 "easy to imagine" is up there with verbal tropes like "Clearly" and "obviously" -I think we'd be better off without them, like we would be without an appendix. "I think", "I suspect", "I intuit" are more honest, because they're closer to how Eliezer suggests we use "magic" as a placeholder. Those acknowledge that bias could be playing a large role, whereas I think "easy to imagine", etc. obfuscate that. “easy to imagine” is up there with verbal tropes like “Clearly” and “obviously” -I think we’d be better off without them, like we would be without an appendix.

“I think”, “I suspect”, “I intuit” are more honest, because they’re closer to how Eliezer suggests we use “magic” as a placeholder. Those acknowledge that bias could be playing a large role, whereas I think “easy to imagine”, etc. obfuscate that.

]]> By: James Bach http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416037 James Bach Sun, 02 Sep 2007 19:16:23 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416037 The availability bias, yes, and also other biases touch on this, such as representativeness bias ("an effect probably looks like its cause"), post hoc ergo propter hoc ("this must be happening because that other thing which happened right before it"), prominence bias (something that grabs our attention or touches our emotions is given more weight as evidence), and assimilation bias ("that new evidence just confirms what I already believe, especially if I kind of squint and hold it sideways..."). Also, the Occam's Razor heuristic seems like an example of easy to imagine. To be easy to imagine is to be matchable to a model already held in our minds. It's not easy for some people to imagine that I could pour water into a glass, turn the glass over, and nothing pours out. But when I show them how Aquagel works (it turns water to a wax-like gel in a few seconds), suddenly it becomes easy to imagine. -- James The availability bias, yes, and also other biases touch on this, such as representativeness bias (“an effect probably looks like its cause”), post hoc ergo propter hoc (“this must be happening because that other thing which happened right before it”), prominence bias (something that grabs our attention or touches our emotions is given more weight as evidence), and assimilation bias (“that new evidence just confirms what I already believe, especially if I kind of squint and hold it sideways…”).

Also, the Occam’s Razor heuristic seems like an example of easy to imagine.

To be easy to imagine is to be matchable to a model already held in our minds. It’s not easy for some people to imagine that I could pour water into a glass, turn the glass over, and nothing pours out. But when I show them how Aquagel works (it turns water to a wax-like gel in a few seconds), suddenly it becomes easy to imagine.

– James

]]> By: Robin Hanson http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416036 Robin Hanson Sun, 02 Sep 2007 18:57:34 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416036 Roy, a heuristic is not necessarily a bias. I was asking if there is a bias here. Roy, a heuristic is not necessarily a bias. I was asking if there is a bias here.

]]> By: Roy Haddad http://www.overcomingbias.com/2007/09/what-evidence-e.html#comment-416035 Roy Haddad Sun, 02 Sep 2007 16:37:36 +0000 http://prod.ob.trike.com.au/2007/09/what-evidence-ease-of-imagination.html#comment-416035 I learned about this one in Psychology 101. <a href="http://en.wikipedia.org/wiki/Availability_heuristic" rel="nofollow">Availability bias</a>: <i>The availability heuristic is a rule of thumb, heuristic, or cognitive bias, where people base their prediction of the frequency of an event or the proportion within a population based on how easily an example can be brought to mind.</i> I learned about this one in Psychology 101.

Availability bias:

The availability heuristic is a rule of thumb, heuristic, or cognitive bias, where people base their prediction of the frequency of an event or the proportion within a population based on how easily an example can be brought to mind.

]]>