[extropy-chat] Bayes, crackpots and psi

[Damien Broderick]

At 11:26 PM 12/19/2004 -0800, Hal wrote:

>Another possibility is that they don't really disagree as much as
>they seem to.  It could be that if they were forced to come up with a
>percentage estimate for the probability that psychic powers exist, they
>wouldn't be that far off.  Maybe it is merely a matter of perspective,
>the glass being half full or half empty.

I doubt it. Eliezer knows the glass is empty, and that in fact there's no 
glass. I still don't know. The evidence is slippery; some of it is largely 
masked (apparently) by security classification, but enough is public to 
make it seem worth pursuing; some is almost as preposterous as Cramer TI 
quantum theory and nonlocal entanglement.

The penumbra of silly beliefs that accretes around psi fanciers turns me 
off big time, but that will blow away once the mechanism is found and some 
moderately reliable application is developed.

On the other hand, I've been expecting such breakthroughs for some decades 
now, ever since statistically sophisticated specialists began exploring the 
paranormal with ever-more carefully designed electronic devices and 
computerized record-keeping, yet the phenomena do keep shifting and 
skidding away. But *not*, it seems, in the fashion of misunderstood 
outliers or of out and out liars.

Meanwhile, people I really have no motive to distrust *except* for these 
declarations affirm that they've witnessed macroscale anomalies apparently 
associated with unmediated intention. If it ever happens to me, I'll 
conclude that psi is genuine. By contrast, if I have a flabbergasting 
mystical experience of the John Wright sort, I'll probably assume that my 
brain went on the fritz in a most agreeable or Rilkean-terrifying way.

I do concur with Eliezer that real evidence will need to be p < 0.001, or 
less. This is increasingly a view shared by paranormalists like Professor 
Ertel.

Damien Broderick