Chris Lakin

On Beliefs

What’s useful may not be true.

Often, the beliefs and tools seem to be true conflict with the beliefs that are actually useful. In which case, we should often take note of our confusion, and use these useful beliefs anyway.

I’ll begin this essay perhaps controversially, and then I’ll extrapolate. (I’m not religious.)


I have found no sound scientific arguments for the existence of God. But this doesn’t mean that believing in a religion is useless.

Religion isn’t about facts and objective reality, and it shouldn’t be. Instead, religion is about beliefs and utility.

For some people at least, religion can provide benefits:

  1. it can provide a philosophy—one which has also been refined over thousands of years—for how to live; and
  2. it can provide a filtered social group of people also living by the same philosophy. And these people are easier to trust (at a base rate) than those outside of the religion.

But it doesn’t matter at all whether religion is ‘strictly true’. Instead, what matters is whether religion helps someone live a better life.

Beliefs like religion aren’t about truth—they’re about utility.



Indigenous Americans have a tradition of processing grain in a particular way before eating. Anyone that doesn’t do this and eats grains for a long period of time gets sick.

It wasn’t known until the twentieth century that their processing method prevents what would otherwise create a niacin deficiency. But this was a tradition— they didn’t have to know why it worked, they just had to do it.

And supposedly, original colonial settlers were taught this process by Indigenous Americans, but failed to keep the tradition. The resulting sickness was the most widespread nutritional deficiency in American history. The cause was not widely known until the 20th century.


How is it that, despite tremendous pressure towards forgetting and towards simplification, complex traditions and stories persist across millennia?

Because the people who believed the current surviving stories (there’s survivorship bias here) out-reproduced those who didn’t, furthering both the people and the meme they carried.


We don’t entirely know how anesthesia works. Yet it’s been used medically for a long time. And anesthesia definitely isn’t the treatment in modern medicine that’s like this.


Strictly, all models are imprecise.

Strictly, all maps are incorrect.

For traditions, there is a natural selection process that filters the traditions that have lasted long enough to be observed today—and for this reason, we should trust them even if we don’t know why they work.

Yeah, some practices turn out wrong (ex), but some turn out right (ex).

So while complete scientific data is of course preferable, consideration of how long a belief has persisted—despite the pressures forgetting and simplification—can be valuable.

We shouldn’t limit ourselves to only acting on directives from what’s studied only scientifically. This allows us to leverage wisdom beyond our current capabilities.

Limits of this

Whenever you use a tool which you don’t understand, add it to a list of “tools I use but do not understand”. It can be easy to forget (and then never question) the source of beliefs!

Moreover, take note of when a tool or belief is useful, and in what situations it isn’t. We don’t have to understand why the heuristic fails, either, just when it does.

(This is especially important when building something. Take note of every assumption or logical-rounding that was crucial in the development of your creation.)


In life we’re not necessarily optimizing for truth— we’re optimizing for value.

But in practice, what seems to be true sometimes conflicts with what tools are actually useful. In which case, we should take note of our confusion, but use the tool anyway.

It’s an artificial limitation to limit yourself merely to what seems to be true, and exclude the beliefs that are merely useful and not necessarily true. (And this, itself, is a useful belief to have!)


Ideated 2020 July, posted 2020 September 18, last updated 2021 March 11.

Get notified when new essays are posted.