Be careful. The infinite sum of epsilons over an infinite domain and over the time domain can and usually does converge to a finite value in real life.
Also, you can avoid having sex with a HIV-positive person. You can probably not avoid driving. So the fact that a risk may be low is not sufficient to take such risk if you can hedge against it.
It is because everyone dies relatively quickly that you can ignore ultra low risks. Suppose your lifespan was potentially 1 trillion years, how low would your risk tolerance need to be?
Meanwhile doing some once a day with a 1 in 1,000,000,000 chance of possibility killing you in 40 years just isn’t worth considering.
> Suppose your lifespan was potentially 1 trillion years, how low would your risk tolerance need to be?
Depends what your quality lifespan is. If I get sick of everything, that time isn't worth much.
There's also a "time-value" of life.
Cool stuff I can do today may be worth a lot more to me than things I can do 500B years from now. Especially since the person that would be doing them is someone I cannot identify with at all.
I don’t think most peoples 25 year old self can really identify what their 85 year old self will be like. Yet saving for retirement isn’t just about them some time in the future it’s about making you feel better now.
> I don’t think most peoples 25 year old self can really identify what their 85 year old self will be like.
Yes, and I believe this is a reason why people make choices that adversely affect their 50 year old self. Who's that guy in the future going to be? Why should I care about making it better for him?
You could say the same thing about people drinking heavily only to wake up the next day with a massive hangover. So it isn’t the differences between them that’s causing people to screw over their future self.
Also, you can avoid having sex with a HIV-positive person. You can probably not avoid driving. So the fact that a risk may be low is not sufficient to take such risk if you can hedge against it.