[ExI] Fermi question, was is a FTL drive a dream . . .

[Kelly Anderson]

On Thu, Dec 22, 2011 at 6:14 AM, Eugen Leitl <eugen at leitl.org> wrote:
> On Thu, Dec 22, 2011 at 03:56:50AM -0700, Kelly Anderson wrote:
>
>> Probably... but if you grant that such a simulation is possible...
>> then it seems far more likely that we are one of the Gazillion
>
> You're not listening. By using the word "likely" you're obviously using
> a probabilistic estimate. But you're outside of statistics'
> scope of applicability. It works over there. It doesn't work here.

Why not? Really. Why not?

> As you're not observing the entire ensemble but just do a self-measurement
> the probability of self-detection is unity regardless of whether there
> is 10^0, 10^1, 10^12 or 10^30 instances of self-observation. If the outcome
> is always unity regardless of what's in the exponent, what does this say to you?
> That the outcome of the measurement is not a function of the number of
> observations, as long you can't cross-correlate them (omniscient
> observer -- not you). You're perfectly myopic.
> Self-measurements are perfectly biased. They're no good.
> You do know that you do exist. Cogito, ergo sum still applies.

I understand that cogito, ergo sum still applies. But I'm really not
following the logic. Perhaps it is too late. I know I can't tell
whether I'm in a simulation, but that is independent of whether or not
I really AM in a simulation.

> You cannot distinguish the individual cases with the information
> you have. This applies across space and across time. It applies
> both to probability of sentient life in the universe or how many
> people have lived when. No simulation argument for you. Also, no pony.

I wasn't trying for a pony. I'd actually be incrementally happier if I
weren't in a simulation... LOL.

> This isn't hard. Why have people such trouble getting it?

Sorry, but it seems hard to me. Would you mind trying again? Maybe go
slow and use smaller words... :-)

-Kelly