I am very excited to announce my new book: If No One Builds It, Everyone Dies.

My Story

On an early morning late last year, while eating my daily Huel squared,^[1] I came across an ad for a new book entitled If Anyone Builds It, Everyone Dies. Despite what struck me as a rather modest thesis, the book was receiving an extraordinary amount of publicity (and no small amount of controversy). I was stunned. I had not realized that the existence of mortality was a matter of serious debate.

Some might believe that every Effective Altruist has firsthand knowledge of mortality due to all of the people they’ve [personally, literally, and not hyperbolically] killed, but it turns out this is not the case. Apparently, even the proponents of mortality’s existence are unwilling to make the strong claim that “people die.” Instead, they must argue the weaker claim that “people die in worlds where artificial superintelligence exists.”

This is where I come in. I will fill the crucial gap in mortality existence theory.

You might be wondering, why am I well suited for this problem? Do I have some sort of relevant domain expertise? No. But I have good^[2] epistemics and a passing interest in philosophy, which makes me qualified to solve any problem the world may throw at me.

My Argument

Epistemic status: Ready to throw ad hominems and/or punches at anyone who doubts me

Here is the core of my argument:

• Suppose that death isn’t real.
  • That means that extinction isn’t real, so the human population will grow until my new longtermist intervention looks cost-effective without bound.^[3]
  • As people, anyone who is not an EA will be genetically similar to EAs, making them “EA adjacent.”
  • The definition of EA is so broad and poorly agreed upon that surely it wouldn’t hurt to stretch it a bit further, right?
  • Therefore, even with zero community building efforts, the number of (immortal) EAs will grow without bound.
• After this post goes live, I am going to drop a quarter into a wishing well. I will wish that every immortal EA spends many years talking about things they like without using caveats.
  • The probability that wishing wells work is p(well) > 0.
  • The probability of my plan working is 25 * p(well) > 0, because I’m using a quarter instead of a penny.
  • I will calibrate my wish to last for 1/p(well) years.
  • That means the expected number of years is 25 * p(well) * 1/p(well)  = 25 years.
• For an EA, talking about things they like without caveats is excruciating pain, and they will do anything in their power to avoid it, including building a time machine to stop me from publishing this post.
  • The probability that this works is P(time machine) > 0.
• Because the number of EAs grows without bound, one of the EAs will eventually build a time machine and stop me from posting this.
• If you are reading this, nobody built a time machine to stop me, which means that death must be real.

My argument is perfect. However, some people have started mandating that we need “evidence” for our claims before we redirect a large portion of the movements’ resources. Based on historical precedent, I think this must be a new rule? Regardless, we are in luck...

Some groundbreaking research from Noduh et al. has demonstrated that every single person to ever live (up until the last 100 years or so) has died. They have expanded their model to say that everyone will die, not just those born a long time ago.^[4] This work was widely influential at the time of publication and was expected to win the Nobel prize. Unfortunately, the prize can not be awarded posthumously.

I’ll go into more detail about my argument, the evidence, counterarguments, and implications^[5] in my book. In this post, I will simply cover a few Frequently Asked Questions.

FAQs

What if technology allows us to stop aging and dying?

As a good student of EA, I know that the only technology that actually exists is Artificial General Intelligence. However, I am 100% confident that If Anyone Builds It, Everyone Dies because the authors of that book said so, and there is no chance they made a mistake.

How does this affect farmed animal welfare work?

Because death is inevitable, we are all approaching it at all times. All animal welfare work should thus be recategorized as humane slaughter. This recategorization is not simply semantics but a crucial definitional change, which is different from semantics in ways I will not elaborate on here.

Does this mean saving lives through GiveWell’s top charities is of zero moral value?

Yes.

Can I give you feedback?

No, but you can leave your name in this form along with nothing else.

Are you going to die too?

1. ^{^}

   I freeze Huel into cubes and put them in liquid Huel so that my consumption is 2x as effective.

2. ^{^}

   I have not ever conducted any sort of test of my epistemics throughout the course of my entire life. However, everyone in my social circle agrees with my ideas, so they must be really good.

3. ^{^}

   Assuming Ambitious Impact eventually stops incubating cost-effective family planning charities.

^{^}

The prior work of Adam & Eve (4000 BCE), lacked the statistical power to generalize beyond the original sample size of two.

^{^}

Such as, you should give a lot of money to me specifically.

Show all footnotes