<html><head></head><body><div>Here is a part from a paper I am working on:</div><div><br></div><div><br></div><div><div><div>"Another relevant property of a civilization is its temporal discounting. How much is the far future worth relative to the present? There are several reasons to suspect advanced civilizations have very long time horizons.</div><div><br></div><div>In a dangerous or uncertain environment it is rational to rapidly discount the value of a future good since survival to that point is not guaranteed. However, mature expanding civilizations have likely reduced their existential risks to a minimum level and would have little reason to discount strongly (individual members, if short-lived, may of course have high discount rates). More generally, the uncertainty of the future will be lower and this also implies lower discount rates. </div><div><br></div><div>It also appears likely that a sufficiently advanced civilization could regulate its ``mental speed'', either by existing as software running on hardware with a variable clock speed, or by simply hibernating in a stable state for a period. If this is true, then the value of something after a period of pause/hibernation would be determined not by the chronological external time, but how much time the civilization would subjectively experience in waiting for it. Changes in mental speed can hence make temporally remote goods more valuable if the observer can pause until they become available and there is no alternative cost for other goods.</div><div><br></div><div>This is linked to a reduction of opportunity costs: advanced civilizations have mainly ``seen it all'' in the present universe and do not gain much more information utility from hanging around in the early era\footnote{Some exploration, automated or ``manned'', might of course still occur during the aestivation period, to be reported to the main part of the civilization at its end.}</div><div><br></div><div>There are also arguments that future goods should not be discounted in cases like this. What really counts is fundamental goods (such as well-being or value) rather than commodities; while discounting prices of commodities makes economic sense it may not make sense to discount value itself \cite{Broome}. </div><div><br></div><div>This is why even a civilization with some temporal discounting can find it rational to pause in order to gain a huge reward in the far future. If the subjective experience is an instant astronomical multiplication of goods (with little risk) it is rational to make the jump. "</div></div><div><br></div><div><br></div><div>This is used in a larger argument that advanced civilizations might aestivate until a late cosmological era, in order to use the low temperature to do more computation. This in itself does not solve Fermi, but I also argue that in this scenario there would have to be caretaker systems to monitor things and protect resources, and these may have good reasons to be 'quiet'.</div></div><div><br></div>The above reasoning suggests that even very fast civilizations may do long-term stuff, even if the aestivation idea itself is wrong. <div><br><br>Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University</div></body></html>