Hacker News Reader

Comment by boxed | Hacker News Reader

The initial problem is to define, quantify, and guard against evil. Imo this problem is basically ignored.

I have a simple definition I call the Grand Unified Theory of Evil:

Evil = delusion * conviction * power

Delusion = how factually incorrect your idea is compared with base reality

Conviction = how little you will try to find flaws with the idea, how little you will test it, how likely you are to implement the idea at 100% of the universe if you had total power

Power = Your ability to push your idea into the world

This model isn't a tool to calculate evil. It's a tool to reason about how to avoid evil. First you need to be not wrong. Secondly you need to be damn UNSURE about what you know and try it at small scale. And lastly you need to have just enough power to test the idea, and show that it works to others if it does, but not any more.

View on HN

How about this:

Idea: we can grab this land with our troops (assume this is 100% based in reality)

Delusion: 0%

Conviction: idea was checked, is aboslutely correct

Power: 100%

=> Evil = 0 * 100 * 100 = 0

I think the categorical imperative is a way better tool to recognise evil.

The delusion is that this is good for your country. That you missed this delusion and then were totally convinced of your own cleverness even though you missed such an enormous flaw.. well that's kindof the point.

> The initial problem is to define, quantify, and guard against evil.

Not really. I mean "guard against evil" loosely interpreted includes all of life so it's hard to say.

But spending a bunch of time defining, quantifying what is evil, arguing about the specifics on the internet, that is deep in the weeds exactly where they want you.

Most people have a good enough intuition for it when they see it up close. The menace is in how well we've obscured a lot of the true evil behind layers of misdirection and diffusion. Focus more on developing your innate sense of wrong so you can ably intuit what is actually happening. That is much more practically valuable than getting bogged down in constructing theories of what evil is and what is evil and how much.

> That is much more practically valuable than getting bogged down in constructing theories of what evil is and what is evil and how much.

I wrote "This model isn't a tool to calculate evil."

Spotting evil is all well and good, but what about when you CAN'T spot it? What then? You STILL need to guard against it. Say you're Mao, now you believe in your ideas. But you should not be convinced, because being doubtful is a guard against evil. You should not use your full power, because this is again a guard against evil.

If Mao had followed this, he would not have killed tens of millions of his own people.

I think you're missing good old corruption from your model but otherwise ya.

I worked for a company in their project management area and it was straightforward and not fraught with moral judgement.

Later I got promoted (Crying emoji) to a new role and suddenly everything I dealt with seemed filled with interests and corruption. This jump corresponded with the budgets of the projects and programs I dealt with.

Same org, different parts of the org. Bad promotion :(

Corruption is captured fairly well by this. It's a delusion in itself: by doing something for yourself short term you will in fact hurt yourself, your children and everyone you love in the long run.

Delusion in this case runs quite deep.

Reply on news.ycombinator.com