I don’t believe in making up credence numbers, especially for an audience full of people who can’t tell whether or not to trust my perspective.
The very reason that I am here is to get EAs sufficiently engaged with the right material so that they can approach these sorts of questions themselves, and then we can have a real discussion.
How do I define the potential global catastrophe and people killed? I think I’ve already answered that in my original post: nuclear war. Of course, there is debate about whether a nuclear war would be a global or local catastrophic risk, though HaydnBelfield’s original post is subject to this same question.
Local ethnic cleansing is much more likely than nuclear war.
If you are looking to understand my priors, then imagine the breakup of Yugoslavia, but in the entire EU, not just the Balkans. And the EU is much more strategically relevant to the world’s nuclear powers than Yugoslavia. The EU also contains over a million Russians, and Putin would be motivated to protect them if the EU started to collapse.
an audience full of people who can’t tell whether or not to trust my perspective.
Statements like “There is a growing risk that European countries will fall into civil war” are very implausible to many folks here. So if you want people to take you seriously, you should at least show us that you sincerely believe this, by being willing to turn those statements into testable predictions. Your refusal to do this is part of the reason some of us don’t trust your perspective.
I don’t think that’s fair; experts in all sorts of domains refrain from making probabilistic predictions despite having strong and legitimate opinions. The idea of using subjective credences to measure beliefs is a unique practice which is uncommon in circles outside EA.
My perspective is implausible to people here because they are trapped in a media bubble. If you want to know my confidence level, it can be inferred by the way I discuss this subject. If people don’t trust what I’m saying because of my conclusions, then how does it help for me to quantify my confidence level by pulling numbers out of the air? My confidence level will clash with their worldview, which will be used as an excuse to dismiss the content of my argument regardless of its merits.
I realize that asking for predictions or bets is a common norm here, but I don’t think it helps us in this case due to the level of inferential distance. There is no point in trying to trade predictions based on certain trends or get into the nuances of forecasting when people don’t agree on those trends, or whether predictions have or haven’t been falsified. This is exactly why Popper is not the last word and there are many objections to falsifiability.
In theory, me giving predictions could be helpful for people to figure out whether my model is correct (at least the people who are being honest), but then we would have to wait and see whether those predictions come true, and be able to agree on whether they have come true. If I predicted X number of mass rapes or terrorist attacks in 2017, and it came true, would people finally admit that maybe there was a problem?
Your reaction to the world is going to be very slow if you cannot update on anything until someone’s predictions come true months or years later.
I believe that there is already significant evidence that there is something wrong, and by giving you predictions about the future, I would be underselling my case about the past and the present.
So then how should this audience evaluate my perspective?
They could check the sources I’ve provided, and see whether they provide novel and important knowledge about the world, and the trajectory of Europe. Even if someone doesn’t agree with everything I say or fully trust my calibration, they would be doing themselves a disservice if they dismiss the novel material I have provided here, instead of investigating it further.
They could attempt to extrapolate predictions from my model and then see whether those predictions match the future world.
Rather than asking me for predictions about the future, they could look at the past record of predictions of people with similar models to mine. For example, is the Rotherham mass child rape scandal more consistent with the predictions of multiculturalists and open borders advocates, or is it more consistent with a perspective that is citizenist, nativist, restrictionist, or right-wing?
For another example of predictions, see Enoch Powell’s Rivers of Blood speech from 1968 (original audio), for which he was crucified as a “racist,” even though it’s quite moderate. If you look at some of Powell’s predictions (native Brits becoming strangers in their own country, neighborhoods changed beyond recognition, crime, discrimination against natives under the guise of anti-discrimination, natives socially punished for speaking up), I think it’s quite clear that if Powell was still alive today, he would feel vindicated. As for “rivers of blood,” that hasn’t happened in England yet, but it has happened in France during the Bataclan attack (warning: graphic).
Powell was capable of forseeing 2016 in 1968. In contrast, people in the media bubble cannot postdict events that have already happened (e.g. Rotherham, Brexit, Trump).
I don’t believe in making up credence numbers, especially for an audience full of people who can’t tell whether or not to trust my perspective.
The very reason that I am here is to get EAs sufficiently engaged with the right material so that they can approach these sorts of questions themselves, and then we can have a real discussion.
How do I define the potential global catastrophe and people killed? I think I’ve already answered that in my original post: nuclear war. Of course, there is debate about whether a nuclear war would be a global or local catastrophic risk, though HaydnBelfield’s original post is subject to this same question.
Local ethnic cleansing is much more likely than nuclear war.
If you are looking to understand my priors, then imagine the breakup of Yugoslavia, but in the entire EU, not just the Balkans. And the EU is much more strategically relevant to the world’s nuclear powers than Yugoslavia. The EU also contains over a million Russians, and Putin would be motivated to protect them if the EU started to collapse.
Short explanation of the breakup of Yugoslavia
The BBC’s excellent documentary
Comparison of Yugoslavia to European democracies
Statements like “There is a growing risk that European countries will fall into civil war” are very implausible to many folks here. So if you want people to take you seriously, you should at least show us that you sincerely believe this, by being willing to turn those statements into testable predictions. Your refusal to do this is part of the reason some of us don’t trust your perspective.
I don’t think that’s fair; experts in all sorts of domains refrain from making probabilistic predictions despite having strong and legitimate opinions. The idea of using subjective credences to measure beliefs is a unique practice which is uncommon in circles outside EA.
My perspective is implausible to people here because they are trapped in a media bubble. If you want to know my confidence level, it can be inferred by the way I discuss this subject. If people don’t trust what I’m saying because of my conclusions, then how does it help for me to quantify my confidence level by pulling numbers out of the air? My confidence level will clash with their worldview, which will be used as an excuse to dismiss the content of my argument regardless of its merits.
I realize that asking for predictions or bets is a common norm here, but I don’t think it helps us in this case due to the level of inferential distance. There is no point in trying to trade predictions based on certain trends or get into the nuances of forecasting when people don’t agree on those trends, or whether predictions have or haven’t been falsified. This is exactly why Popper is not the last word and there are many objections to falsifiability.
In theory, me giving predictions could be helpful for people to figure out whether my model is correct (at least the people who are being honest), but then we would have to wait and see whether those predictions come true, and be able to agree on whether they have come true. If I predicted X number of mass rapes or terrorist attacks in 2017, and it came true, would people finally admit that maybe there was a problem?
Your reaction to the world is going to be very slow if you cannot update on anything until someone’s predictions come true months or years later.
I believe that there is already significant evidence that there is something wrong, and by giving you predictions about the future, I would be underselling my case about the past and the present.
So then how should this audience evaluate my perspective?
They could check the sources I’ve provided, and see whether they provide novel and important knowledge about the world, and the trajectory of Europe. Even if someone doesn’t agree with everything I say or fully trust my calibration, they would be doing themselves a disservice if they dismiss the novel material I have provided here, instead of investigating it further.
They could attempt to extrapolate predictions from my model and then see whether those predictions match the future world.
Rather than asking me for predictions about the future, they could look at the past record of predictions of people with similar models to mine. For example, is the Rotherham mass child rape scandal more consistent with the predictions of multiculturalists and open borders advocates, or is it more consistent with a perspective that is citizenist, nativist, restrictionist, or right-wing?
For another example of predictions, see Enoch Powell’s Rivers of Blood speech from 1968 (original audio), for which he was crucified as a “racist,” even though it’s quite moderate. If you look at some of Powell’s predictions (native Brits becoming strangers in their own country, neighborhoods changed beyond recognition, crime, discrimination against natives under the guise of anti-discrimination, natives socially punished for speaking up), I think it’s quite clear that if Powell was still alive today, he would feel vindicated. As for “rivers of blood,” that hasn’t happened in England yet, but it has happened in France during the Bataclan attack (warning: graphic).
Powell was capable of forseeing 2016 in 1968. In contrast, people in the media bubble cannot postdict events that have already happened (e.g. Rotherham, Brexit, Trump).
Muslim jihadists have done a truck attack in Berlin and assassinated the Russian ambassador to Turkey. Me two days ago: