You don't need to define good and bad, instead you focus on better and worse and the metrics used to measure them. Now the goal is to maximize "better" and minimize "worse." You may recognize this as the essentials utilitarianism. The advantage utilitarianism has is it can be applied algorithmically without passion or emotion - in other words, by AI.
Utilitarianism leads to controversial outcomes, but every decision is defensible.
For each thing T, that T is defensible under at least one ethical framework.
Teaching an optimiser AI any of those frameworks, or even any preference ordering or combination function within Utiliarianism (because value({T, T}) doesn't have to equal 2 * value({T})), will lead to it optimising what you said, without necessarily limiting that to situations anything close to the training distribution.
To put it another way: if you run an AB test on a social media site and it observes that people are more likely to engage with content that makes them angry, then tell it to boost engagement "because socialising is always good, obviously" then it will get your users as angry as possible and suddenly you get Buddhists going off and committing surprise genocide before anyone tells you something has gone wrong.
I would argue this has been known for decades and is in fact the origin of one of the earliest memes in computer science: To err is human, but to really mess things up requires a computer!
Opportunistic, and morally equivalant to the destruction of the other 72 Japanese cities, and the bombing of European cities by both the Allies and the Axis.
I’m saying that it is possible to learn how to optimise for “better” or “worse” decisions without being trained on an opinionated dataset of every single event that has happened. An “intelligence” could exist without necessarily answering your nukes question.
You’re bringing alignment to a capability discussion.
How do you suppose we do this? It makes decisions based on numbers alone ?
Personally, I don't think there is a right decision, it was just a decision which had an outcome, for some people it was ab absolutely fucking devastating decision.
This is where I think we have blind spots when developing these systems, I don't think there is a "right" or best answer here. Just some answer.
Utilitarianism leads to controversial outcomes, but every decision is defensible.