Re: An Interesting Poll

Anders Sandberg (
18 Jun 1999 16:16:13 +0200

Doug Bailey <> writes:

> I often have to defend my lifestyle habits in the context of my
> transhumanist ideas. I exercise, watch my diet, wear my seat belt,
> no sunbathing, etc. A few people have asked me why care about my
> cholesterol intake is or whether I get skin cancer when I believe
> that MNT, uploading, etc. may be developed in the near future.
> My response is that while I believe such technologies are virtually
> certain to be developed at some point, I don't know when that will
> be. By maximizing my lifespan, I increase the probability that
> I will be able to benefit from these advance technologies.

Exactly my view too. Of course, living a healthy life doesn't mean being an ascetic, and one has to strike a balance between pleasure and health sometimes (e.g. I use too much caffeine and enjoy the occasional huge meal, but it is worth it).

What I find a bit disturbing is the "Nanotech Santa will fix it all"-meme. It is often not based on any serious estimation of when sufficiently good nanotechnology will arrive and in what health state one will be at that point, and to outsiders it sounds far too religious to inspire confidence in transhumanism. Everybody has to determine their lifestyle for themselves, but it better be a rational choice (even if it involves a large risk).

One can view it as an optimization problem. You can select v(t), the amount of "vices" (i.e. harmful but pleasant stuff) you indulge in over time. That will of course accumulate until L(T(t),V(t)), the maximal lifespan possible given a certain technological level T(t) and total amount of vice (V(t) = integral of v(t) from 0 to t).

So the total amount of pleasure achieved from vices (I don't count all the healthy pleasures) is a function P(V(t)) of V(t), and for a lifespan up to t_max you get P(V(t_max)) where t_max is the first solution of L(T(t_max),V(t_max)) = t_max.

If we make some assumptions that T(t) increases like c_1 exp(c_2 t) and L(t)=75+ exp (c1 t) - V(t) (where c1 is a constant of how fast tech advances lifespan), then the equation for lifespan becomes 75 + exp(c1 t_max) - integral_0^t_max v(t) = t_max If we further assume people hold a constant pleasure level we get 75 + exp(c1 t_max) - v*t_max = t_max which gives

(1+v) t_max = 75+exp(c1 t_max)

Even if you live a perfectly "virtuous" life (v=0), you will die eventually if c1 is too small. In this case, we can assume c1 to be close to zero and the total pleasure over life P(t_max) = v * 76/(1+v), which increases with v - life fast and die young.

But if c1 is large enough, then the above equation may lack a solution. That means that for sufficiently low v, you can life forever and the total pleasure over life will diverge. In fact, in this case the optimal strategy would be to first choose a v just large enough for the curve (1+v)*t to be tangent to 75+exp(c1 t), and after this point increase v exponentially. A more cautious person might chose a somewhat lower v.

So if we are believers in the high value of c1, we should be relatively careful until lifespans start to expand greatly, and then go for heavy duty hedonism. If we believe c1 is not large enough, well, après nous le déluge...

OK, this is a game with equations and not to be taken that seriously. But the field of hedonic planning is clearly well worth investigating, if only just for fun.

Anders Sandberg                                      Towards Ascension!                  
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y