[ExI] Self-Driving Cars Must Make Ethical Decisions
connor_flexman at brown.edu
Wed Jul 29 18:32:29 UTC 2015
On Wed, Jul 29, 2015 at 11:49 AM, William Flynn Wallace <foozler83 at gmail.com
> Rationality is about drawing correct inferences from limited,
> confusing, contradictory, or maliciously doctored facts.
> -- Scott Alexander
> I'd say this is sheer dumb luck. How can one be rational when the 'facts'
> are false, or the data aren't complete, or are ambiguous or even completely
> biased by doctoring?
> bill w
You can read the answer at
point isn't that you literally arrive to a correct conclusion based on
actually doctored data. The point is that you take a bunch of conflicting
nutrition studies, some of which may be doctored, some of which have crappy
technique, some of which are noise, and pull out of that data the best
thing you can: a conclusion that may be relatively weak and have large
confidence intervals, but is nonetheless not what many people get out of
it, which is "X is clearly correct, the other studies are bunk, I am done
with this topic". Darwin being rational enough to piece together natural
selection from extremely limited evidence is also an example Alexander uses.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the extropy-chat