<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
On 2016-05-16 04:16, Rafal Smigrodzki wrote:<br>
<blockquote
cite="mid:CAAc1gFgzx0phW3D7-tO3tzZMVz_6skABBME9m95D3zR9S8_5Ug@mail.gmail.com"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote"><br>
<div>Is the technogenesis likelihood of 10e-15 per solar
mass a reasonable estimate?</div>
</div>
</div>
</div>
</blockquote>
<br>
Our view is that is a reasonable estimate. Although it is a fair bit
higher than what our priors suggest - one civilization per multiple
observable universe-volume is closer to our mean. The variance of
the prior is *big* (I actually had to do numerical voodoo for some
of the calculations on my computer, since precision breaks down for
absurdly small numbers).<br>
<br>
The thing is, people tend to assume that if you see one instance of
something out of N objects, the probability of it happening should
be about 1/N. But this misses the observer selection effect: if you
have an absurdly low probability but it has to happen somewhere
(because it results in you), then the relevant class may be far
larger than N. So if the technogenesis probability is 10^-1000 there
will still be lots of observers in an infinite or very large
universe, but they will all be very distant. <br>
<br>
<pre class="moz-signature" cols="72">--
Anders Sandberg
Future of Humanity Institute
Oxford Martin School
Oxford University</pre>
</body>
</html>