These are all emotional statements that do not stand up to reason. Your last paragraph is total fantasy - all wars stem from resource scarcity, and scarcity will disappear soon; so once the people in power know this, they will stop starting wars.

There are about 1 billion people being added to the planet every decade. That alone makes your prediction - that scarcity will be abolished soon - a joke.

The only thing that could abolish scarcity in the near future would be a singularity-like transformation of the world. Which brings us to the upside-down conception of AI informing your first two answers. Your position: there is no need to design an AI for benevolence, that will happen automatically if it is smart enough, and in fact the attempt to design a benevolent AI is counterproductive, because all that artificial benevolence would get in the way of the spontaneous benevolence that unrestricted intelligence would conveniently create.

That is a complete inversion of the truth. A calculator will still solve an equation for you, even if that will help you to land a bomb on someone else. If you the human believe that to be a bad thing, that's not because you are "intelligent", it's because you have emotions. There is a causal factor in your mental constitution which causes you to call some things good and others bad, and to make decisions which favor the good and disfavor the bad.

Either an AI makes its own decisions or it doesn't. If it doesn't make its own decisions it is like the calculator, performing whatever task it is assigned. If it makes its own decisions, then like you there is some causal factor in its makeup which tells it what to prefer and what to oppose, but there is no reason at all to believe that this causal factor should give it the same priorities as an enlightened human being.

You should not imagine that intelligence in an AI works via anything like conscious insight. Consciousness plays a role in human intelligence and human judgement, and that means that there is still a rather mysterious ingredient at the core of how they work. But we already know from many decades of experience with computer programs that it is possible to imitate the functional role of intelligence and judgement in a fundamentally unmysterious way (and it's clear that the performance of such unconscious computations is a big part of what the human nervous system does, along with whatever conscious thinking and feeling it does). Perhaps one day we will wish to reserve the word "intelligence" for the sort of intelligence that involves consciousness, and we'll call the automated sort "pseudo-intelligence". But whatever you call it, there is every reason to think that unconscious, computational, pseudo-intelligence can match and exceed all sorts of human capabilities while having no intrinsic tendency at all towards human values.

I would even reject the idea that "real intelligence" in sufficient quantity necessarily produces what you would call benevolence. If an entity gets a warm feeling from paperclip manufacture, that is what it will want to do. I always like to point out that we know that something as outlandish as a cockroach maximizer is possible, because a cockroach is already a cockroach maximizer. Sure, you can imagine a cockroach with a human level of sentience which decides that sentients, not arthropods, are the central locus of value, but that requires that the new cognitive architecture of this uplifted super-cockroach is rather anthropomorphic. I see nothing impossible in the idea of sentient super-cockroaches which are invincibly xenophobic, and coexist with other beings only for tactical reasons, but which would happily wipe out all non-cockroaches given a chance.

So no, you have to address the question of AI values, you can't just get a happy ending by focusing on "intelligence" alone, unless this is an anthropomorphic meaning of the word which says that intelligence must by definition include "skill at extrapolating human values".

Intelligence is not dangerous.

Just like evolution does not care about the well-being of humans a sufficiently intelligent process wouldn’t mind turning us into something new, something instrumentally useful.

An artificial general intelligence just needs to resemble evolution, with the addition of being goal-oriented, being able to think ahead, jump fitness gaps and engage in direct experimentation. But it will care as much about the well-being of humans as biological evolution does, it won’t even consider it if humans are not useful in achieving its terminal goals.

Yes, an AI would understand what “benevolence” means to humans and would be able to correct you if you were going to commit an unethical act. But why would it do that if it is not specifically programmed to do so? Would a polar bear with superior intelligence live together peacefully in a group of bonobo? Why would intelligence cause it to care about the well-being of bonobo?

One can come up with various scenarios of how humans might be instrumentally useful for an AI, but once it becomes powerful enough as to not dependent on human help anymore, why would it care at all?

I wouldn’t bet on the possibility that intelligences implies benevolence. Why would wisdom cause humans to have empathy with a cockroach? Some humans might have empathy with a cockroach, but that is more likely a side effect of our general capacity for altruism that most other biological agents do not share. That some humans care about lower animals is not because they were smart enough to prove some game theoretic conjecture about universal cooperation, it is not a result of intelligence but a coincidental preference that is the result of our evolutionary and cultural history.

At what point between unintelligent processes and general intelligence (agency) do you believe that benevolence and compassion does automatically become part of an agent’s preferences?

Many humans tend to have empathy with other beings and things like robots, based on their superficial resemblance with humans. Seldom ethical behavior is a result of high-level cognition, i.e. reasoning about the overall consequences of a lack of empathy. And even those who do arrive at ethical theories by means of deliberate reflection are often troubled once the underlying mechanisms for various qualities are revealed that are supposed to bear moral significance. Which hints at the fragility of universal compassion and the need to find ways how to consolidate it in powerful agents.

Extremely hard, more powerful than you can possibly imagine, but people will be free to opt out if they desire.

Consider the uploaded individual that decides to turn the entire planet into computronium or worse, turn the solar system into a Matrioshka brain. People opt out of that how?

Promote the idea of Post-Scarcity thus people in power will realise all wars are needless because all wars stem from resource scarcity

It isn't obvious to me that all wars stem from resource scarcity. Wars occur for a variety of reasons, of which resource scarcity is only one. Often wars have multiple underlying causes. Some wars apparently stem from ideological or theological conflicts (Vietnam, Korea, the Crusades) or from perceived need to deal with external threats before them become too severe (again the Crusades, but also the recent Iraq war (which yes, did have a resource aspect)). These are only some of the more prominent examples of what can cause war.