On the difference between x-risks and x-risk factors
I suspect there isn’t much of a meaningful difference between “x-risks” and “x-risk factors,” for two reasons:
We can treat them the same in terms of probability theory. For example, if X is an “x-risk” and Y is a “risk factor” for X, then Pr(X∣Y)>Pr(X). But we can also say that Pr(Y∣X)>Pr(Y), because both statements are equivalent to Pr(X,Y)>Pr(X)Pr(Y). We can similarly speak of the total probability of an x-risk factor because of the law of total probability (e.g. Pr(Y)=Pr(Y∣X1)+Pr(Y∣X2)+…) like we can with an x-risk.
Concretely, something can be both an x-risk and a risk factor. Climate change is often cited as an example: it could cause an existential catastrophe directly by making all of Earth unable to support complex societies, or indirectly by increasing humanity’s vulnerability to other risks. Pandemics might also be an example, as a pandemic could either directly cause the collapse of civilization or expose humanity to other risks.
I think the difference is that x-risks are events that directly cause an existential catastrophe, such as extinction or civilizational collapse, whereas x-risk factors are events that don’t have a direct causal pathway to x-catastrophe. But it’s possible that pretty much all x-risks are risk factors and vice versa. For example, suppose that humanity is already decimated by a global pandemic, and then a war causes the permanent collapse of civilization. We usually think of pandemics as risks and wars as risk factors, but in this scenario, the war is the x-risk because it happened last… right?
One way to think about x-risks that avoids this problem is that x-risks can have both direct and indirect causal pathways to x-catastrophe.
I think your comment (and particularly the first point) has much more to do with the difficulty of defining causality than with x-risks.
It seems natural to talk about force causing the mass to accelerate: when I push a sofa, I cause it to start moving. but Newtonian mechanics can’t capture casualty basically because the equality sign in →F=m→a lacks direction. Similarly, it’s hard to capture causality in probability spaces.
Following Pearl, I come to think that causality arises from manipulator/manipulated distinction.
So I think it’s fair to speak about factors only with relation to some framing:
If you are focusing on bio policy, you are likely to take great-power conflict as an external factor.
Similarly, if you are focusing on preventing nuclear war between India and Pakistan, you are likely to take bioterrorism as an external factor.
Usually, there are multiple external factors in your x-risk modeling. The most salient and undesirable are important enough to care about them (and give them a name).
Calling bio-risks an x-factor makes sense formally; but doesn’t make sense pragmatically because bio-risks are very salient (in our community) on their own because they are a canonical x-risk. So for me, part of the difference is that I started to care about x-risks first; and that I started to care about x-risk factors because of their relationship to x-risk.
On the difference between x-risks and x-risk factors
I suspect there isn’t much of a meaningful difference between “x-risks” and “x-risk factors,” for two reasons:
We can treat them the same in terms of probability theory. For example, if X is an “x-risk” and Y is a “risk factor” for X, then Pr(X∣Y)>Pr(X). But we can also say that Pr(Y∣X)>Pr(Y), because both statements are equivalent to Pr(X,Y)>Pr(X)Pr(Y). We can similarly speak of the total probability of an x-risk factor because of the law of total probability (e.g. Pr(Y)=Pr(Y∣X1)+Pr(Y∣X2)+…) like we can with an x-risk.
Concretely, something can be both an x-risk and a risk factor. Climate change is often cited as an example: it could cause an existential catastrophe directly by making all of Earth unable to support complex societies, or indirectly by increasing humanity’s vulnerability to other risks. Pandemics might also be an example, as a pandemic could either directly cause the collapse of civilization or expose humanity to other risks.
I think the difference is that x-risks are events that directly cause an existential catastrophe, such as extinction or civilizational collapse, whereas x-risk factors are events that don’t have a direct causal pathway to x-catastrophe. But it’s possible that pretty much all x-risks are risk factors and vice versa. For example, suppose that humanity is already decimated by a global pandemic, and then a war causes the permanent collapse of civilization. We usually think of pandemics as risks and wars as risk factors, but in this scenario, the war is the x-risk because it happened last… right?
One way to think about x-risks that avoids this problem is that x-risks can have both direct and indirect causal pathways to x-catastrophe.
I think your comment (and particularly the first point) has much more to do with the difficulty of defining causality than with x-risks.
It seems natural to talk about force causing the mass to accelerate: when I push a sofa, I cause it to start moving. but Newtonian mechanics can’t capture casualty basically because the equality sign in →F=m→a lacks direction. Similarly, it’s hard to capture causality in probability spaces.
Following Pearl, I come to think that causality arises from manipulator/manipulated distinction.
So I think it’s fair to speak about factors only with relation to some framing:
If you are focusing on bio policy, you are likely to take great-power conflict as an external factor.
Similarly, if you are focusing on preventing nuclear war between India and Pakistan, you are likely to take bioterrorism as an external factor.
Usually, there are multiple external factors in your x-risk modeling. The most salient and undesirable are important enough to care about them (and give them a name).
Calling bio-risks an x-factor makes sense formally; but doesn’t make sense pragmatically because bio-risks are very salient (in our community) on their own because they are a canonical x-risk. So for me, part of the difference is that I started to care about x-risks first; and that I started to care about x-risk factors because of their relationship to x-risk.