Ok, but the point of the question is to try to arrive at true beliefs. So imagine forgetting that I’d asked the question. What does your model of the world, which says that simulation is silly, say for the probability that a major establishment scientist who is in no way a transhumanist, believes that we could be in a simulation? If it assigns too low a probability, maybe you should consider assigning some probability to alternative models?
I would not be at all surprised. No speculation is too silly to have been seriously propounded by some philosopher or other, and lofty state gives no immunity to silliness.
[ETA: And of course, I’m talking about ideas that I’ve judged silly despite their being seriously propounded by (some) folks here on LessWrong that I think are really smart, and after reading a whole lot of their stuff before arriving at that conclusion. So one more smart person, however prestigious, isn’t going to make a difference.]
But you changed it to “could be”. Sure, could be, but that’s like Descartes’ speculations about a trickster demon faking all our sensations. It’s unfalsifiable unless you deliberately put something into the speculation to let the denizens discover their true state, but at that point you’re just writing speculative fiction.
But if this person is arguing that we probably are in a simulation, then no, I just tune that out.
So the bottom line of your reasoning is quite safe from any evidential threats?
In one sense, yes, but in another sense....yes.
First sense: I have a high probability for speculations on whether we are living in a simulation (or any of the other ideas I dismiss) not being worth my while outside of entertaining fictions. As a result, evidence to the contrary is unlikely to reach my notice, and even if it does, it has a lot of convincing to do. In that sense, it is as safe as any confidently held belief is from evidential threats.
Second sense: Any evidential threats at all? Now we’re into unproductive navel-gazing. If, as a proper Bayesian, I make sure that my probabilities are never quite equal to 1, and therefore answer that my belief must be threatened by some sort of evidence, the next thing is you’ll ask what that evidence might be. But why should anyone have to be able to answer that question? If I choose to question some idea I have, then, yes, I must decide what possible observations I might make that would tell either way. This may be a non-trivial task. (Perhaps for reasons relating to the small world/large world controversy in Bayesian reasoning, but I haven’t worked that out.) But I have other things to do—I cannot be questioning everything all the time. The “silly” ideas are the ones I can’t be bothered spending any time on at all even if people are talking about them on my favorite blog, and if that means I miss getting in on the ground floor of the revelation of the age, well, that’s the risk I accept in hitting the Ignore button.
So in practice, yes, my bottom line on this matter (which was not written down in advance, but reached after having read a bunch of stuff of the sort I don’t read any more) is indeed quite safe. I don’t see anything wrong with that.
Besides that, I am always suspicious of this question, “what would convince you that you are wrong?” It’s the sort of thing that creationists arguing against evolution end up saying. After vigorously debating the evidence and making no headway, the creationist asks, “well, what would convince you?”, to which the answer is that to start with, all of the evidence that has just been gone over would have to go away. But in the creationist’s mind, the greater their failure to convince someone, the greater the proof that they’re right and the other wrong. “Consider it possible that you are mistaken” is the sound of a firing pin clicking on an empty chamber.
“what would convince you that you are wrong?” It’s the sort of thing that creationists arguing against evolution
But a proponent of evolution can easily answer this, for example if they went to the fossil record and found it showed that all and only existing creatures’ skeletons appeared 6000 years ago, and that radiocarbon dating showed that the earth was 6000 years old.
Ok, but the point of the question is to try to arrive at true beliefs. So imagine forgetting that I’d asked the question. What does your model of the world, which says that simulation is silly, say for the probability that a major establishment scientist who is in no way a transhumanist, believes that we could be in a simulation? If it assigns too low a probability, maybe you should consider assigning some probability to alternative models?
I would not be at all surprised. No speculation is too silly to have been seriously propounded by some philosopher or other, and lofty state gives no immunity to silliness.
[ETA: And of course, I’m talking about ideas that I’ve judged silly despite their being seriously propounded by (some) folks here on LessWrong that I think are really smart, and after reading a whole lot of their stuff before arriving at that conclusion. So one more smart person, however prestigious, isn’t going to make a difference.]
But you changed it to “could be”. Sure, could be, but that’s like Descartes’ speculations about a trickster demon faking all our sensations. It’s unfalsifiable unless you deliberately put something into the speculation to let the denizens discover their true state, but at that point you’re just writing speculative fiction.
But if this person is arguing that we probably are in a simulation, then no, I just tune that out.
So the bottom line of your reasoning is quite safe from any evidential threats?
In one sense, yes, but in another sense....yes.
First sense: I have a high probability for speculations on whether we are living in a simulation (or any of the other ideas I dismiss) not being worth my while outside of entertaining fictions. As a result, evidence to the contrary is unlikely to reach my notice, and even if it does, it has a lot of convincing to do. In that sense, it is as safe as any confidently held belief is from evidential threats.
Second sense: Any evidential threats at all? Now we’re into unproductive navel-gazing. If, as a proper Bayesian, I make sure that my probabilities are never quite equal to 1, and therefore answer that my belief must be threatened by some sort of evidence, the next thing is you’ll ask what that evidence might be. But why should anyone have to be able to answer that question? If I choose to question some idea I have, then, yes, I must decide what possible observations I might make that would tell either way. This may be a non-trivial task. (Perhaps for reasons relating to the small world/large world controversy in Bayesian reasoning, but I haven’t worked that out.) But I have other things to do—I cannot be questioning everything all the time. The “silly” ideas are the ones I can’t be bothered spending any time on at all even if people are talking about them on my favorite blog, and if that means I miss getting in on the ground floor of the revelation of the age, well, that’s the risk I accept in hitting the Ignore button.
So in practice, yes, my bottom line on this matter (which was not written down in advance, but reached after having read a bunch of stuff of the sort I don’t read any more) is indeed quite safe. I don’t see anything wrong with that.
Besides that, I am always suspicious of this question, “what would convince you that you are wrong?” It’s the sort of thing that creationists arguing against evolution end up saying. After vigorously debating the evidence and making no headway, the creationist asks, “well, what would convince you?”, to which the answer is that to start with, all of the evidence that has just been gone over would have to go away. But in the creationist’s mind, the greater their failure to convince someone, the greater the proof that they’re right and the other wrong. “Consider it possible that you are mistaken” is the sound of a firing pin clicking on an empty chamber.
But a proponent of evolution can easily answer this, for example if they went to the fossil record and found it showed that all and only existing creatures’ skeletons appeared 6000 years ago, and that radiocarbon dating showed that the earth was 6000 years old.
The creationist generally puts his universal question after having unsuccessfully argued that the fossil record and radiocarbon dating support him.