On the Forum, invalid meta-arguments* are pretty common, such as âpeople make logic mistakes so you might have tooâ, rather than actually identifying the weaknesses in an argument.
Here are two claims Iâd very much agree with:
Itâs often best to focus on object-level arguments rather than meta-level arguments, especially arguments alleging bias
One reason for that is that the meta-level arguments will often apply to a similar extent to a huge number of claims/âpeople. E.g., a huge number of claims might be influenced substantially by confirmation bias.
But you say invalid meta-arguments, and then give the example âpeople make logic mistakes so you might have tooâ. That example seems perfectly valid, just often not very useful.
And Iâd also say that that example meta-argument could sometimes be useful. In particular, if someone seems extremely confident about something based on a particular chain of logical steps, it can be useful to remind them that there have been people in similar situations in the past whoâve been wrong (though also some whoâve been right). Theyâre often wrong for reasons âoutside their modelâ, so this person not seeing any reason theyâd be wrong doesnât provide extremely strong evidence that theyâre not.
It would be invalid to say, based on that alone, âYouâre probably wrongâ, but saying theyâre plausibly wrong seems both true and potentially useful.
(Also, isnât your comment primarily meta-arguments of a somewhat similar nature to âpeople make logic mistakes so you might have tooâ? I guess your comment is intended to be a bit closer to a specific reference class forecast type argument?)
Thereâs also a lot of pseudo-superforcasting, like âI have 80% confidence in thisâ, without any evidence backing up those credences.
Describing that as pseudo-superforecasting feels unnecessarily pejorative. I think such people are just forecasting /â providing estimates. They may indeed be inspired by Tetlockâs work or other work with superforecasters, but that doesnât mean theyâre necessarily trying to claim their estimates use the same methodologies or deserve the same weight as superforecastersâ estimates. (I do think there are potential downsides of using explicit probabilities, but I think each potential downside is debatable, and there are also potential upsides, and using seemingly pejorative terms probably doesnât help.)
Effective altruists have centred around some ideas that are correct (longtermism, moral uncertainty, etc.)
Did you mean âsome ideas that are probably correct and very importantâ? If so, Iâd agree. But I wouldnât want to imply longtermism (and to a lesser extent moral uncertainty) are simply âcorrectâ (rather than âquite likelyâ or âwhat we should act based on, given (meta)moral uncertainty and expected value reasoning.
but outside of that, Iâd say weâre just as wrong as anyone else.
Iâd disagree with that. I definitely donât think EAs are perfect, but they do seem above-average in their tendency to have true beliefs and update appropriately on evidence, across a wide range of domains.
But you say invalid meta-arguments, and then give the example âpeople make logic mistakes so you might have tooâ. That example seems perfectly valid, just often not very useful.
My definition of an invalid argument contains âarguments that donât reliably differentiate between good and bad argumentsâ. â1+1=2âł is also a correct statement, but that doesnât make it a valid response to any given argument. Arguments need to have relevancy. I dunno, I could be using âinvalidâ incorrectly here.
And Iâd also say that that example meta-argument could sometimes be useful.
Yes, if someone believed that having a logical argument is a guarantee, and theyâve never had one of their logical arguments have a surprising flaw, it would be valid to point that out. Thatâs fair. But (as you seem to agree with) the best way to do this is to actually point to the flaw in the specific argument theyâve made. And since most people who are proficient with logic already know that logic arguments can be unsound, itâs not useful to reiterate that point to them.
Also, isnât your comment primarily meta-arguments of a somewhat similar nature to âpeople make logic mistakes so you might have tooâ?
It is, but as I said, âSome meta-arguments are validâ. (I can describe how I delineate between valid and invalid meta-arguments if you wish.)
Describing that as pseudo-superforecasting feels unnecessarily pejorative.
Ah sorry, I didnât mean to offend. If they were superforecasters, their credence alone would update mine. But theyâre probably not, so I donât understand why they give their credence without a supporting argument.
Did you mean âsome ideas that are probably correct and very importantâ?
The set of things I give 100% credence is very, very small (i.e. claims that are true even if Iâm a brain in a vat). I could say âThereâs probably a table in front of meâ, which is technically more correct than saying that there definitely is, but it doesnât seem valuable to qualify every statement like that.
Why am I confident in moral uncertainty? People do update their morality over time, which means that they were wrong at some point (i.e. there is demonstrably moral uncertainty), or the definition of âcorrectâ changes and nobody is ever wrong. I think ânobody is ever wrongâ is highly unlikely, especially because you can point to logical contradictions in peopleâs moral beliefs (not just unintuitive conclusions). At that point, itâs not worth mentioning the uncertainty I have.
I definitely donât think EAs are perfect, but they do seem above-average in their tendency to have true beliefs and update appropriately on evidence, across a wide range of domains.
Yeah, Iâm too focused on the errors. Iâll concede your point: Some proportion of EAs are here because they correctly evaluated the arguments. So theyâre going to bump up the average, even outside of EAâs central ideas. My reference classes here were all the groups that have correct central ideas, and yet are very poor reasoners outside of their domain. My experience with EAs is too limited to support my initial claim.
Why am I confident in moral uncertainty? People do update their morality over time [...]
Oh, when you said âEffective altruists have centred around some ideas that are correct (longtermism, moral uncertainty, etc.)â, I assumed (perhaps mistakenly) that by âmoral uncertaintyâ you meant something vaguely like the idea that âWe should take moral uncertainty seriously, and think carefully about how best to handle it, rather than necessarily just going with whatever moral theory currently seems best to us.â
So not just the idea that we canât be certain about morality (which Iâd be happy to say is just âcorrectâ), but also the idea that that fact should change our behaviour is substantial ways. I think that both of those ideas are surprisingly rare outside of EA, but the latter one is rarer, and perhaps more distinctive to EA (though not unique to EA, as there are some non-EA philosophers whoâve done relevant work in that area).
On my âinside-viewâ, the idea that we should âtake moral uncertainty seriouslyâ also seems extremely hard to contest. But I move a little away from such confidence, and probably wouldnât simply call it âcorrectâ, due to the fact that most non-EAs donât seem to explicitly endorse something clearly like that idea. (Though maybe they endorse somewhat similar ideas in practice, even just via ideas like âagree to disagreeâ.)
Here are two claims Iâd very much agree with:
Itâs often best to focus on object-level arguments rather than meta-level arguments, especially arguments alleging bias
One reason for that is that the meta-level arguments will often apply to a similar extent to a huge number of claims/âpeople. E.g., a huge number of claims might be influenced substantially by confirmation bias.
(Here are two relevant posts.)
Is that what you meant?
But you say invalid meta-arguments, and then give the example âpeople make logic mistakes so you might have tooâ. That example seems perfectly valid, just often not very useful.
And Iâd also say that that example meta-argument could sometimes be useful. In particular, if someone seems extremely confident about something based on a particular chain of logical steps, it can be useful to remind them that there have been people in similar situations in the past whoâve been wrong (though also some whoâve been right). Theyâre often wrong for reasons âoutside their modelâ, so this person not seeing any reason theyâd be wrong doesnât provide extremely strong evidence that theyâre not.
It would be invalid to say, based on that alone, âYouâre probably wrongâ, but saying theyâre plausibly wrong seems both true and potentially useful.
(Also, isnât your comment primarily meta-arguments of a somewhat similar nature to âpeople make logic mistakes so you might have tooâ? I guess your comment is intended to be a bit closer to a specific reference class forecast type argument?)
Describing that as pseudo-superforecasting feels unnecessarily pejorative. I think such people are just forecasting /â providing estimates. They may indeed be inspired by Tetlockâs work or other work with superforecasters, but that doesnât mean theyâre necessarily trying to claim their estimates use the same methodologies or deserve the same weight as superforecastersâ estimates. (I do think there are potential downsides of using explicit probabilities, but I think each potential downside is debatable, and there are also potential upsides, and using seemingly pejorative terms probably doesnât help.)
Did you mean âsome ideas that are probably correct and very importantâ? If so, Iâd agree. But I wouldnât want to imply longtermism (and to a lesser extent moral uncertainty) are simply âcorrectâ (rather than âquite likelyâ or âwhat we should act based on, given (meta)moral uncertainty and expected value reasoning.
Iâd disagree with that. I definitely donât think EAs are perfect, but they do seem above-average in their tendency to have true beliefs and update appropriately on evidence, across a wide range of domains.
My definition of an invalid argument contains âarguments that donât reliably differentiate between good and bad argumentsâ. â1+1=2âł is also a correct statement, but that doesnât make it a valid response to any given argument. Arguments need to have relevancy. I dunno, I could be using âinvalidâ incorrectly here.
Yes, if someone believed that having a logical argument is a guarantee, and theyâve never had one of their logical arguments have a surprising flaw, it would be valid to point that out. Thatâs fair. But (as you seem to agree with) the best way to do this is to actually point to the flaw in the specific argument theyâve made. And since most people who are proficient with logic already know that logic arguments can be unsound, itâs not useful to reiterate that point to them.
It is, but as I said, âSome meta-arguments are validâ. (I can describe how I delineate between valid and invalid meta-arguments if you wish.)
Ah sorry, I didnât mean to offend. If they were superforecasters, their credence alone would update mine. But theyâre probably not, so I donât understand why they give their credence without a supporting argument.
The set of things I give 100% credence is very, very small (i.e. claims that are true even if Iâm a brain in a vat). I could say âThereâs probably a table in front of meâ, which is technically more correct than saying that there definitely is, but it doesnât seem valuable to qualify every statement like that.
Why am I confident in moral uncertainty? People do update their morality over time, which means that they were wrong at some point (i.e. there is demonstrably moral uncertainty), or the definition of âcorrectâ changes and nobody is ever wrong. I think ânobody is ever wrongâ is highly unlikely, especially because you can point to logical contradictions in peopleâs moral beliefs (not just unintuitive conclusions). At that point, itâs not worth mentioning the uncertainty I have.
Yeah, Iâm too focused on the errors. Iâll concede your point: Some proportion of EAs are here because they correctly evaluated the arguments. So theyâre going to bump up the average, even outside of EAâs central ideas. My reference classes here were all the groups that have correct central ideas, and yet are very poor reasoners outside of their domain. My experience with EAs is too limited to support my initial claim.
Oh, when you said âEffective altruists have centred around some ideas that are correct (longtermism, moral uncertainty, etc.)â, I assumed (perhaps mistakenly) that by âmoral uncertaintyâ you meant something vaguely like the idea that âWe should take moral uncertainty seriously, and think carefully about how best to handle it, rather than necessarily just going with whatever moral theory currently seems best to us.â
So not just the idea that we canât be certain about morality (which Iâd be happy to say is just âcorrectâ), but also the idea that that fact should change our behaviour is substantial ways. I think that both of those ideas are surprisingly rare outside of EA, but the latter one is rarer, and perhaps more distinctive to EA (though not unique to EA, as there are some non-EA philosophers whoâve done relevant work in that area).
On my âinside-viewâ, the idea that we should âtake moral uncertainty seriouslyâ also seems extremely hard to contest. But I move a little away from such confidence, and probably wouldnât simply call it âcorrectâ, due to the fact that most non-EAs donât seem to explicitly endorse something clearly like that idea. (Though maybe they endorse somewhat similar ideas in practice, even just via ideas like âagree to disagreeâ.)