- Have we really forsaken natural selection?
- We don’t trade with ants
- Pacing: inexplicably good
- Let’s think about slowing down AI
- Counterarguments to the basic AI risk case
- A game of mattering
- A game of mattering
- Podcasts on surveys, slower AI, AI arguments
- What do ML researchers think about AI in 2022?
- Why do people avoid vaccination?
Morality is subjective preference, but it can be objectively wrong
Posted on June 18, 2009
People are often unwilling to think of ethics as their own preferences, rather than demands from something more transcendent. For instance it’s normal to claim that one really wants to make one choice, but it’s only ethical to make the other. My feelings agree, but my thoughts don’t. If I follow something I call ethics, that demonstrates that I want to. It’s not a physical law. So what’s the difference?
Just that. Ethics is a preference for fulfilling preferences attributed to some other source. Popular external sources of values include Gods, nature, other people, transcendent moral truth, group norms, and leaders. If I prefer for your house not to burn down I will turn on the hose. If I think it’s moral to stop your house burning down I will turn off the hose if I find out that you want to burn it down to collect insurance money. I care about your values, not the house.
One demonstration that having an external source is important for ethics is the fact that invented ethical systems (such as, ‘playing video games is virtuous’) seem illegitimate and cheaty. Crazy seeming practices can be ordained by religion and culture, but if you decide independently that it’s only ethical to eat cereal on Thursdays and most will feel you are missing the point and some marbles.
While ethics is a matter of choice then, it implies the existence of your preferred outside source of values. This means it can be wrong. The outside source of values might not exist, or might not have values. This is why evidence about evolution can influence whether a person likes gays marrying, despite it being an apparent value judgement.
This means moral intuitions aren’t as useful as they seem for information about how to be moral. Gut reactions are handy for working out what you like, but if you find that you like serving someone else’s purposes there is factual information about whether they exist or care to take into account. We have better ways to deal with facts than our emotional responses in most realms, so why not use the same here?
The only things that exist and care that I know of are other people and animals. Gods and transcendent values don’t exist, and society as a whole and the environment don’t care, as far as I know. So if I want to be ethical, preference utilitarianism (caring about other people’s preferences) is my only option. Of course I could prefer not to be ethical at all. And I could prefer to follow what pass for other moral rules; being honest, protesting interference in the environment, keeping my dress long. But if these things benefit only my feeling of righteousness, I must admit they are no different to normal personal preferences. If you want to be ethical, these are probably not what you are looking for any more than ‘it’s virtuous to play video games’ is.