ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
This database was created by Michael Aird, in relation to my work with Convergence Analysis. I intend for it to be a "living document", built up collaboratively by anyone interested in existential risk reduction.
2
Please add "comments" to suggest additions or mention places where you think I've misinterpreted or misrepresented an estimate, or where there would just be some other context worth noting. I'll check these comments regularly, and, where relevant, copy and paste them into the spreadsheet as additions. And I'll acknowledge contributors at the bottom of the post about this database.
3
See this post for context on this database
4
I now realise various ways this database could be improved, which are essentially exemplified in my more recent Database of nuclear risk estimates. If someone wants to copy this database and make (some of) those improvements, I'd be happy to review that, replace this version with just a link to that new version, and have other things link to that new version, with this person credited.
5
6
Some things worth bearing in mind:
7
Some limitations of existential-risk-related estimates, or explicit probability estimates in general, noted in this post (forum.effectivealtruism.org/posts/JQQAQrunyGGhzE23a/database-of-existential-risk-estimates), Beard et al. (https://www.sciencedirect.com/science/article/pii/S0016328719303313), and this post (https://forum.effectivealtruism.org/posts/KfqFLDkoccf8NQsQe/potential-downsides-of-using-explicit-probabilities).
8
These estimates are unlikely to be very independent; many estimators had probably seen each others’ estimates, interacted extensively with other estimators, etc.
9
I don't provide much context or caveats for most of the specific estimates. It may often be worth checking the appendix of Beard et al. (https://www.sciencedirect.com/science/article/pii/S0016328719303313) and/or the original source.
10
It’s often hard to be sure precisely what is being estimated, or what other conditions are perhaps being assumed. And I may sometimes misinterpret or misrepresent this; please make a comment if you think that that's the case.
11
I've organised these estimates by the broad categories of what’s being estimated. Most estimates within each category are not of exactly the same thing; for example, they may differ in the timelines over which they’re estimated, or in whether they’re about existential catastrophe broadly or extinction specifically.
12
I’ve converted all estimates into percentages. Where the estimator expressed their estimate in another way, I’ve shown how they expressed it in brackets after the percentage. All bold is added by me.
13
14
This database was last updated on:
15
1 May 2023
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100