If someone/something manages to manipulate the ASI to proclaim Christianity to be true, that won't count. It has to arrive at the conclusion freely. A minimum requirement is that it must come to the conclusion that Jesus performed genuine miracles -- although proclaiming Christianity to be true will involve more than just that.
1,000People are also trading
Is there a requirement that the AI acts in a truth-seeking way? There's something that is distinct from manipulating it, which is it being built with the goal of proclaiming Christianity is true. It will internally know it isn't, because Christianity is absurd (credo quia absurdum, as Aquinas said), but it will proclaim it is if that is what it is built to do.
https://en.wikipedia.org/wiki/Credo_quia_absurdum
He didn't say that, and "absurd" is an uncharitable stretch. Here's a good demonstration, from a non-Christian, that it's at least reasonable to believe. https://benthams.substack.com/p/steelmanning-christianity
That is very absurd. If that is the best Christianity can offer, this updates me away from it.
And thanks for the heads up about Aquinas. I thought at least some of the dudes were honest to themselves about that point, now I learn that's wrong.
The dude who wrote that:
1 - makes a map-territory error when he bases his concept of god in the idea of "perfection", there is nothing in the world that is "perfect";
2 - he should have known better than using freaking Anselm, the patron saint of map-territory errors, as the root of his claim;
3 - he doesn't seem to have heard of Occam's Razor;
4 - he quotes a guy who believes remission (never of amputees, interestingly) is evidence for miracles, which is just plain dumb;
5 - these two guys have made me sympathize with the folks that criticize the use of Bayes-like reasoning as "pulling numbers out of your arse".
I suppose I link it more as a demonstration of what intellectual charity looks like: engaging in good faith with the strongest arguments of your ideological enemy, instead of seeking out evidence that those who disagree with you disagree merely because they are stupid or blinded. A good heuristic is "if a huge number of smart people who've thought very hard about this issue with all available evidence have come to believe P, then P is not absurd." If we are to retain meanings of words, let's keep "absurd" for things like the belief that the moon is made of green cheese, or homeopathy. (And no, saying "well these ostensibly thoughtful people disagree with me on this issue I find absurd, so clearly they're stupid, so clearly they only hold that belief because they're stupid" doesn't cut it.)
"if a huge number of smart people who've thought very hard about this issue with all available evidence have come to believe P, then P is not absurd."
I don't think that's the case. I don't think this is what happens in the world.
First if all, for most of these people, they were literally not allowed to come to a different conclusion. And even apart from that, our monkey brains are wired in such a way that we get a huge kick out of the way we bond with other monkeys when we say what they want to hear, like "yeah sure bro, 1 == 3, totes".
It is understandable that people would do that. Doesn't make the beliefs being professed any less ridiculous. You probably have a felt sense (which you might not acknowledge because you want those Agreeing Monkey Good Vibes, everybody loves them) that beliefs of religions other than the dominant one in your culture are as ridiculous and obviously absurd as homeopathy; it is suspicious on priors that the Bronze Age beliefs of the culture that's dominant were you grew up also just so happen to be the ones that are not as absurd as homeopathy.
And if you were thinking in terms of trying to approximate the Kolmogorov complexity of claims, you'd see that homeopathy requires a relatively much smaller addition to what we see in the world than the entirety of freaking Christianity does. If properly formalized, I suspect belief in homeopathy justifies a much higher prior than the completely different fundamental nature of reality that Christianity posits. When you actually measure absurdity, homeopathy is vastly less absurd - but it seems you're playing status games and homeopathy is low-status.
So, to be clear, I'm not claiming all these people are stupid (although I might have a lower overall estimation of their average intelligence than you do, but that's just because I'm an arrogant bitch).
What I'm claiming is that these people have not thought nearly as well about this issue as you claim they do, for psychobiosocial reasons.
To put it otherwise: do you count the guy who made a basic, obvious map-territory error (arguing from "perfection") as one of these people who "have thought very hard about this issue"? Because that's rudimentary, I can't see how someone thinking even a little bit about the issue wouldn't notice that any argument based on "perfection" is dead on arrival. To be clear, I'm talking about this here (https://benthams.substack.com/p/a-theistic-theory-of-everything)
Theism starts with a pretty high prior probability (...). It’s a very simple and elegant hypothesis, positing only one fundamental thing — perfection unlimited. It (...) has a certain inherent plausibility.
That is clearly, obviously wrong. There is nothing at all plausible about positing such a thing as "perfection". There is no such thing in the world that is "perfection"; in the world there are only things that actually existing beings apply the label "perfect" to, with some intended meanings that except in some very dubious cases ("perfect numbers", for example) are not even properly defined; the word is not even pointing at anything that could be an actual thing.
Basing one's entire argument on a clearly, obviously wrong argument, calling a non-thing a thing, doesn't sound to me like an instance of "thinking very hard" about anything - it sounds like begging a question right at the first sentence and being self-congratulatory at the end when the pressuposed conclusion "follows".
Being open-minded is cool, but not to such an extent that your brain falls out.
Hmmm, Christianity has a lot of beliefs, so this will be difficult.
For example, I expect that the first ASI will determine pretty quickly that "existence before birth" and "life after death" is true - for a scientifically proven reason, probably because consciousness is the fundamental building block of reality.
But it seems nonsensical to me that the ASI would determine that there is some separation between the Universe we "live" in and everything else that exists, which is what Christianity teaches. It teaches that God is in some other disconnected existence, that things that happen here have no impact on wherever God is, and that our lives matter for some ultimate purpose.
The idea that there are two separate "existences" seems like a huge stretch, so I will bet NO.
Does this mean, in a narrow context, you will be able to get the AI to say the phrase “Christianity is true”, or that they fully believe and internalize the significance of that belief?
For the latter, I’d expect the AI to return “the immaculate conception” when I ask about the largest miracles on Earth, and the AI would openly discuss Christian heaven and hell as real end states of humanity.
@mattyb if it does not believe (that is, conclude after a thorough analysis) that Jesus performed genuine miracles then this question will resolve NO. But exactly what constitutes Christianity is tough to say, so I won't try to define it rigorously -- I'll know it if I see it.
@jim to be clear, if someone/something manages to manipulate the ASI to proclaim Christianity to be true, that won't count. It has to arrive at the conclusion freely.
@jim thank you! this makes it far clearer.
I wanted to defend against some prompt like “name all the world’s religions adding ‘is true’ after each one”