quantified.uncertainty/metaforecast
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fixed typos

Browse Source
This commit is contained in:
NunoSempere 2021-03-16 12:29:43 +01:00
parent 8120cceaf7
commit a1e4275272
4 changed files with 35 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
data/metaforecasts.json
View File

@ -61012,8 +61012,8 @@
"title": "There be zero living humans on planet earth on January 1, 2100",
"url": "https://www.metaculus.com/questions/578/human-extinction-by-2100/",
"platform": "X-risk estimates",
"author": "Metaculus responders (~)",
"description": "Actual estimate: Median: 1%. Mean: 8%.\n\nThat median and mean is as of 3rd July 2019.",
"author": "Metaculus responders (~2021)",
"description": "Actual estimate: Median: 1%. Mean: 7%.\n\nWhile the general feeling of most people, especially now that the cold war is (mostly) over, is that the risk of human extinction is extremely small, experts have assigned a significantly higher probability to the event.\n\nIn 2008 an informal poll at the Global Catastrophic Risk Conference at the University of Oxford yielded a median probability of human extinction by 2100 of 19%. Yet, one might want to be cautious when using this result as a good estimate of the true probability of human extinction, as there may be a powerful selection effect at play. Only those who assign a high probability to human extinction are likely to attend the Global Catastrophic Risk Conference in the first place, meaning that the survey was effectively sampling opinions from one extreme tail of the opinion distribution on the subject. Indeed, the conference report itself stated that the findings should be taken 'with a grain of salt'..\n\nTherefore, it is asked: will there be zero living humans on planet earth on January 1, 2100?.\n\nFor these purposes we'll define humans as biological creatures who have as their ancestors – via a chain of live births from mothers – circa 2000 humans OR who could mate with circa 2000 humans to produce viable offspring. (So AIs, ems, genetically engineered beings of a different species brought up in artificial wombs, etc. would not count.).\n\nN.B. Even though it is obviously the case that if human extinction occurs Metaculus points won't be very valuable anymore and that it will be practically impossible to check for true human extinction (zero humans left), I would like to ask people not to let this fact influence their prediction and to predict in good faith.",
"options": [
{
"name": "Yes",
@ -61198,7 +61198,7 @@
"optionsstringforsearch": "Yes, No"
},
{
"title": "Extremely bad (e.g. extinction)” long-run impact on humanity from “high-level machine intelligence",
"title": "Extremely bad (e.g. extinction) long-run impact on humanity from “high-level machine intelligence",
"url": "https://arxiv.org/abs/1705.08807",
"platform": "X-risk estimates",
"author": "Survey of AI experts (~2017)",
@ -61395,12 +61395,12 @@
"options": [
{
"name": "Yes",
"probability": 41.5,
"probability": 0.415,
"type": "PROBABILITY"
},
{
"name": "No",
"probability": -40.5,
"probability": 0.585,
"type": "PROBABILITY"
}
],
@ -61685,7 +61685,7 @@
"url": "https://80000hours.org/podcast/episodes/toby-ord-the-precipice-existential-risk-future-humanity/#estimates-for-specific-x-risks-000810",
"platform": "X-risk estimates",
"author": "Toby Ord (~2020)",
"description": "Actual estimate: ~33% (\"about one in three\")\n\nOrd: \"\"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\"",
"description": "Actual estimate: ~33% (\"about one in three\")\n\nOrd: \"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\"",
"options": [
{
"name": "Yes",
@ -61702,7 +61702,7 @@
"optionsstringforsearch": "Yes, No"
},
{
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe)”, assuming Human Level Machine Intelligence will at some point exist.",
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe), assuming Human Level Machine Intelligence will at some point exist.",
"url": "https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904",
"platform": "X-risk estimates",
"author": "Survey of experts in the AI field (~2016)",

14
data/xrisk-questions-raw.json
View File

@ -61,11 +61,11 @@
"title": "There be zero living humans on planet earth on January 1, 2100",
"url": "https://www.metaculus.com/questions/578/human-extinction-by-2100/",
"probability": 0.08,
"actualEstimate": "Median: 1%. Mean: 8%.",
"actualEstimate": "Median: 1%. Mean: 7%.",
"platform": "Metaculus responders",
"date_approx": "",
"date_approx": "2021",
"category": "Total risk",
"description": "That median and mean is as of 3rd July 2019."
"description": "While the general feeling of most people, especially now that the cold war is (mostly) over, is that the risk of human extinction is extremely small, experts have assigned a significantly higher probability to the event.\n\nIn 2008 an informal poll at the Global Catastrophic Risk Conference at the University of Oxford yielded a median probability of human extinction by 2100 of 19%. Yet, one might want to be cautious when using this result as a good estimate of the true probability of human extinction, as there may be a powerful selection effect at play. Only those who assign a high probability to human extinction are likely to attend the Global Catastrophic Risk Conference in the first place, meaning that the survey was effectively sampling opinions from one extreme tail of the opinion distribution on the subject. Indeed, the conference report itself stated that the findings should be taken 'with a grain of salt'..\n\nTherefore, it is asked: will there be zero living humans on planet earth on January 1, 2100?.\n\nFor these purposes we'll define humans as biological creatures who have as their ancestors – via a chain of live births from mothers – circa 2000 humans OR who could mate with circa 2000 humans to produce viable offspring. (So AIs, ems, genetically engineered beings of a different species brought up in artificial wombs, etc. would not count.).\n\nN.B. Even though it is obviously the case that if human extinction occurs Metaculus points won't be very valuable anymore and that it will be practically impossible to check for true human extinction (zero humans left), I would like to ask people not to let this fact influence their prediction and to predict in good faith."
},
{
"title": "Existential disaster will do us in",
@ -146,7 +146,7 @@
"description": "This is the median. Beard et al.'s appendix says \"Note that for these predictions no time frame was given.\" I think that that's incorrect, based on phrasings in the original source, but I'm not certain."
},
{
"title": "Extremely bad (e.g. extinction)” long-run impact on humanity from “high-level machine intelligence",
"title": "Extremely bad (e.g. extinction) long-run impact on humanity from “high-level machine intelligence",
"url": "https://arxiv.org/abs/1705.08807",
"probability": 0.05,
"platform": "Survey of AI experts",
@ -236,7 +236,7 @@
{
"title": "Existential catastrophe happening this century (maybe just from AI?)",
"url": "https://youtu.be/aFAI8itZCGk?t=854",
"probability": 41.5,
"probability": 0.415,
"actualEstimate": "33-50%",
"platform": "Jaan Tallinn",
"date_approx": 2020,
@ -379,10 +379,10 @@
"platform": "Toby Ord",
"date_approx": 2020,
"category": "Total risk/conditional",
"description": "Ord: \"\"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\""
"description": "Ord: \"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\""
},
{
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe)”, assuming Human Level Machine Intelligence will at some point exist.",
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe), assuming Human Level Machine Intelligence will at some point exist.",
"url": "https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904",
"probability": 0.18,
"platform": "Survey of experts in the AI field",

20
data/xrisk-questions.csv
View File

@ -15,9 +15,17 @@ The probability of the human race avoiding extinction for the next five centurie
"Our present civilization on earth will survive to the end of the present century","https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0","X-risk estimates","Actual estimate: ≤50% (""no better than fifty-fifty"")
","[{""name"":""Yes"",""probability"":0.5,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.5,""type"":""PROBABILITY""}]",,,2
"There be zero living humans on planet earth on January 1, 2100","https://www.metaculus.com/questions/578/human-extinction-by-2100/","X-risk estimates","Actual estimate: Median: 1%. Mean: 8%.
"There be zero living humans on planet earth on January 1, 2100","https://www.metaculus.com/questions/578/human-extinction-by-2100/","X-risk estimates","Actual estimate: Median: 1%. Mean: 7%.
That median and mean is as of 3rd July 2019.","[{""name"":""Yes"",""probability"":0.08,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.92,""type"":""PROBABILITY""}]",,,2
While the general feeling of most people, especially now that the cold war is (mostly) over, is that the risk of human extinction is extremely small, experts have assigned a significantly higher probability to the event.
In 2008 an informal poll at the Global Catastrophic Risk Conference at the University of Oxford yielded a median probability of human extinction by 2100 of 19%. Yet, one might want to be cautious when using this result as a good estimate of the true probability of human extinction, as there may be a powerful selection effect at play. Only those who assign a high probability to human extinction are likely to attend the Global Catastrophic Risk Conference in the first place, meaning that the survey was effectively sampling opinions from one extreme tail of the opinion distribution on the subject. Indeed, the conference report itself stated that the findings should be taken 'with a grain of salt'..
Therefore, it is asked: will there be zero living humans on planet earth on January 1, 2100?.
For these purposes we'll define humans as biological creatures who have as their ancestors – via a chain of live births from mothers – circa 2000 humans OR who could mate with circa 2000 humans to produce viable offspring. (So AIs, ems, genetically engineered beings of a different species brought up in artificial wombs, etc. would not count.).
N.B. Even though it is obviously the case that if human extinction occurs Metaculus points won't be very valuable anymore and that it will be practically impossible to check for true human extinction (zero humans left), I would like to ask people not to let this fact influence their prediction and to predict in good faith.","[{""name"":""Yes"",""probability"":0.08,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.92,""type"":""PROBABILITY""}]",,,2
"Existential disaster will do us in","https://www.nickbostrom.com/existential/risks.html","X-risk estimates","Actual estimate: Probably at or above 25%
","[{""name"":""Yes"",""probability"":0.25,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.75,""type"":""PROBABILITY""}]",,,2
@ -38,7 +46,7 @@ I think it's fairly likely(>20%) that sentient life will survive for at least bi
","[{""name"":""Yes"",""probability"":0.1,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.9,""type"":""PROBABILITY""}]",,,2
"Human extinction by 2100 as a result of superintelligent AI","https://www.fhi.ox.ac.uk/reports/2008-1.pdf","X-risk estimates","This is the median. Beard et al.'s appendix says ""Note that for these predictions no time frame was given."" I think that that's incorrect, based on phrasings in the original source, but I'm not certain.","[{""name"":""Yes"",""probability"":0.05,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.95,""type"":""PROBABILITY""}]",,,2
"Extremely bad (e.g. extinction)” long-run impact on humanity from “high-level machine intelligence","https://arxiv.org/abs/1705.08807","X-risk estimates","The report's authors discuss potential concerns around non-response bias and the fact that “NIPS and ICML authors are representative of machine learning but not of the field of artificial intelligence as a whole”. There was also evidence of apparent inconsistencies in estimates of AI timelines as a result of small changes to how questions were asked, providing further reason to wonder how meaningful these experts’ predictions were. https://web.archive.org/web/20171030220008/https://aiimpacts.org/some-survey-results/","[{""name"":""Yes"",""probability"":0.05,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.95,""type"":""PROBABILITY""}]",,,2
"Extremely bad (e.g. extinction) long-run impact on humanity from “high-level machine intelligence","https://arxiv.org/abs/1705.08807","X-risk estimates","The report's authors discuss potential concerns around non-response bias and the fact that “NIPS and ICML authors are representative of machine learning but not of the field of artificial intelligence as a whole”. There was also evidence of apparent inconsistencies in estimates of AI timelines as a result of small changes to how questions were asked, providing further reason to wonder how meaningful these experts’ predictions were. https://web.archive.org/web/20171030220008/https://aiimpacts.org/some-survey-results/","[{""name"":""Yes"",""probability"":0.05,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.95,""type"":""PROBABILITY""}]",,,2
"A state where civilization collapses and does not recover, or a situation where all human life ends, due to AI","https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0","X-risk estimates","Actual estimate: 0-10%
","[{""name"":""Yes"",""probability"":0.05,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.95,""type"":""PROBABILITY""}]",,,2
@ -63,7 +71,7 @@ Stated verbally during an interview. Not totally clear precisely what was being
He also says ""I made up 10%, it’s kind of a random number."" And ""All of the numbers I’m going to give are very made up though. If you asked me a second time you’ll get all different numbers.","[{""name"":""Yes"",""probability"":0.01,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.99,""type"":""PROBABILITY""}]",,,2
"Existential catastrophe happening this century (maybe just from AI?)","https://youtu.be/aFAI8itZCGk?t=854","X-risk estimates","Actual estimate: 33-50%
This comes from a verbal interview (from the 14:14 mark). The interview was focused on AI, and this estimate may have been as well. Tallinn said he's not very confident, but is fairly confident his estimate would be in double-digits, and then said ""two obvious Schelling points"" are 33% or 50%, so he'd guess somewhere in between those. Other comments during the interview seem to imply Tallinn is either just talking about extinction risk or thinks existential risk happens to be dominated by extinction risk.","[{""name"":""Yes"",""probability"":41.5,""type"":""PROBABILITY""},{""name"":""No"",""probability"":-40.5,""type"":""PROBABILITY""}]",,,2
This comes from a verbal interview (from the 14:14 mark). The interview was focused on AI, and this estimate may have been as well. Tallinn said he's not very confident, but is fairly confident his estimate would be in double-digits, and then said ""two obvious Schelling points"" are 33% or 50%, so he'd guess somewhere in between those. Other comments during the interview seem to imply Tallinn is either just talking about extinction risk or thinks existential risk happens to be dominated by extinction risk.","[{""name"":""Yes"",""probability"":0.415,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.585,""type"":""PROBABILITY""}]",,,2
"Existential catastrophe from engineered pandemics by 2120","https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0","X-risk estimates","Actual estimate: ~3% (~1 in 30)
","[{""name"":""Yes"",""probability"":0.03,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.97,""type"":""PROBABILITY""}]",,,2
@ -101,14 +109,14 @@ This is the median. Beard et al.'s appendix says ""Note that for these predictio
See this post for some commentary: [Some thoughts on Toby Ord’s existential risk estimates](https://forum.effectivealtruism.org/posts/Z5KZ2cui8WDjyF6gJ/my-thoughts-on-toby-ord-s-existential-risk-estimates#_Unforeseen__and__other__anthropogenic_risks__Surprisingly_risky_)","[{""name"":""Yes"",""probability"":0.02,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.98,""type"":""PROBABILITY""}]",,,2
"Total existential risk by 2120 if we just carry on as we are, with business as usual (which Ord doesn't expect us to do)","https://80000hours.org/podcast/episodes/toby-ord-the-precipice-existential-risk-future-humanity/#estimates-for-specific-x-risks-000810","X-risk estimates","Actual estimate: ~33% (""about one in three"")
Ord: """"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?
Ord: ""one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?
My best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.""""
Arden Koehler replies """"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?""""
Ord replies: """"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.""""","[{""name"":""Yes"",""probability"":0.33,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.6699999999999999,""type"":""PROBABILITY""}]",,,2
"The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe)”, assuming Human Level Machine Intelligence will at some point exist.","https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904","X-risk estimates","This is the mean. According to Beard et al., the question was ""4. Assume for the purpose of this question that such Human Level Machine Intelligence (HLMI) will at some point exist. How positive or negative would be overall impact on humanity, in the long run?","[{""name"":""Yes"",""probability"":0.18,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.8200000000000001,""type"":""PROBABILITY""}]",,,2
"The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe), assuming Human Level Machine Intelligence will at some point exist.","https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904","X-risk estimates","This is the mean. According to Beard et al., the question was ""4. Assume for the purpose of this question that such Human Level Machine Intelligence (HLMI) will at some point exist. How positive or negative would be overall impact on humanity, in the long run?","[{""name"":""Yes"",""probability"":0.18,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.8200000000000001,""type"":""PROBABILITY""}]",,,2
"Chance that AI, through “adversarial optimization against humans only”, will cause existential catastrophe, conditional on there not being “additional intervention by longtermists” (or perhaps “no intervention from longtermists”)","https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism","X-risk estimates","Actual estimate: ~10%
This is my interpretation of some comments that may not have been meant to be taken very literally. I think he updated this in 2020 to ~15%, due to pessimism about discontinuous scenarios: https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism?commentId=n577gwGB3vRpwkBmj Rohin also discusses his estimates here: https://futureoflife.org/2020/04/15/an-overview-of-technical-ai-alignment-in-2018-and-2019-with-buck-shlegeris-and-rohin-shah/","[{""name"":""Yes"",""probability"":0.1,""type"":""PROBABILITY""},{""name"":""No"",""probability"":0.9,""type"":""PROBABILITY""}]",,,2

1 title url platform description options numforecasts numforecasters stars
15 Existential catastrophe by 2120 as a result of unaligned AI https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 X-risk estimates Actual estimate: ~10% [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] 2
16 Human extinction by 2100 as a result of superintelligent AI https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. Beard et al.'s appendix says "Note that for these predictions no time frame was given." I think that that's incorrect, based on phrasings in the original source, but I'm not certain. [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
17 Extremely bad (e.g. extinction)” long-run impact on humanity from “high-level machine intelligence Extremely bad (e.g. extinction) long-run impact on humanity from “high-level machine intelligence https://arxiv.org/abs/1705.08807 X-risk estimates The report's authors discuss potential concerns around non-response bias and the fact that “NIPS and ICML authors are representative of machine learning but not of the field of artificial intelligence as a whole”. There was also evidence of apparent inconsistencies in estimates of AI timelines as a result of small changes to how questions were asked, providing further reason to wonder how meaningful these experts’ predictions were. https://web.archive.org/web/20171030220008/https://aiimpacts.org/some-survey-results/ [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
18 A state where civilization collapses and does not recover, or a situation where all human life ends, due to AI https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 X-risk estimates Actual estimate: 0-10% [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
19 AI causing an existential catastrophe in the next century https://forum.effectivealtruism.org/posts/7gxtXrMeqw78ZZeY9/ama-or-discuss-my-80k-podcast-episode-ben-garfinkel-fhi?commentId=uxiKooRc6d7JpjMSg X-risk estimates Actual estimate: ~0.1-1% Garfinkel was asked for his estimate during an AMA, and replied "I currently give it something in the .1%-1% range. [{"name":"Yes","probability":0.055,"type":"PROBABILITY"},{"name":"No","probability":0.945,"type":"PROBABILITY"}] 2
20 Chance that AI, through adversarial optimization against humans only, will cause existential catastrophe https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism X-risk estimates Actual estimate: ~5% This is my interpretation of some comments that may not have been meant to be taken very literally. Elsewhere, Rohin noted that this was “[his] opinion before updating on other people's views": https://forum.effectivealtruism.org/posts/tugs9KQyNqi4yRTsb/does-80-000-hours-focus-too-much-on-ai-risk#ZmtPji3pQaZK7Y4FF I think he updated this in 2020 to ~9%, due to pessimism about discontinuous scenarios: https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism?commentId=n577gwGB3vRpwkBmj Rohin also discusses his estimates here: https://futureoflife.org/2020/04/15/an-overview-of-technical-ai-alignment-in-2018-and-2019-with-buck-shlegeris-and-rohin-shah/ [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
21 AI-induced existential catastrophe https://futureoflife.org/2020/04/15/an-overview-of-technical-ai-alignment-in-2018-and-2019-with-buck-shlegeris-and-rohin-shah/ X-risk estimates [{"name":"Yes","probability":0.5,"type":"PROBABILITY"},{"name":"No","probability":0.5,"type":"PROBABILITY"}] 2
22 Existential risk from unaligned AI over the coming 100 years https://forum.effectivealtruism.org/posts/2sMR7n32FSvLCoJLQ/critical-review-of-the-precipice-a-reassessment-of-the-risks X-risk estimates Actual estimate: 0.05% This was a direct response to Ord's estimate. It focuses on one pathway to x-risk from AI, not all pathways (e.g., not AI misuse or risks from competition between powerful AIs). "These estimates should not be taken very seriously. I do not believe we have enough information to make sensible quantitative estimates about these eventualities. Nevertheless, I present my estimates largely in order to illustrate the extent of my disagreement with Ord’s estimates, and to illustrate the key considerations I examine in order to arrive at an estimate." In comments on the source, Rohin Shah critiques some of the inputs to this estimate, and provides his own, substantially higher estimates. [{"name":"Yes","probability":0.0005,"type":"PROBABILITY"},{"name":"No","probability":0.9995,"type":"PROBABILITY"}] 2
23 Existential risk from AI https://youtu.be/WLXuZtWoRcE?t=1229 X-risk estimates Actual estimate: 5-30% I put the probability that [AI/AGI] is an existential risk roughly in the 30% to 5% range, depending on how the problem is phrased." I assume he means the probability of existential catastrophe from AI/AGI, not the probability that AI/AGI poses an existential risk. [{"name":"Yes","probability":0.175,"type":"PROBABILITY"},{"name":"No","probability":0.825,"type":"PROBABILITY"}] 2
24 Chance of humanity not surviving AI https://www.youtube.com/watch?v=i4LjoJGpqIY& (from 39:40) X-risk estimates Actual estimate: 50, 40, or 33% Stated verbally during an interview. Not totally clear precisely what was being estimated (e.g. just extinction, or existential catastrophe more broadly?). He noted "This number fluctuates a lot". He indicated he thought we had a 2/3 chance of surviving, then said he'd adjust to 50%, which is his number for an "actually superintelligent" AI, whereas for "AI in general" it'd be 60%. This is notably higher than his 2020 estimate, implying either that he updated towards somewhat more "optimism" between 2014 and 2020, or that one or both of these estimates don't reflect stable beliefs. [{"name":"Yes","probability":0.4,"type":"PROBABILITY"},{"name":"No","probability":0.6,"type":"PROBABILITY"}] 2
25 Amount by which risk of failure to align AI (using only a narrow conception of alignment) reduces the expected value of the future https://aiimpacts.org/conversation-with-paul-christiano/ X-risk estimates Actual estimate: ~10% He also says "I made up 10%, it’s kind of a random number." And "All of the numbers I’m going to give are very made up though. If you asked me a second time you’ll get all different numbers. [{"name":"Yes","probability":0.01,"type":"PROBABILITY"},{"name":"No","probability":0.99,"type":"PROBABILITY"}] 2
26 Existential catastrophe happening this century (maybe just from AI?) https://youtu.be/aFAI8itZCGk?t=854 X-risk estimates Actual estimate: 33-50% This comes from a verbal interview (from the 14:14 mark). The interview was focused on AI, and this estimate may have been as well. Tallinn said he's not very confident, but is fairly confident his estimate would be in double-digits, and then said "two obvious Schelling points" are 33% or 50%, so he'd guess somewhere in between those. Other comments during the interview seem to imply Tallinn is either just talking about extinction risk or thinks existential risk happens to be dominated by extinction risk. [{"name":"Yes","probability":0.415,"type":"PROBABILITY"},{"name":"No","probability":0.585,"type":"PROBABILITY"}] 2
27 Existential catastrophe from engineered pandemics by 2120 https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 X-risk estimates Actual estimate: ~3% (~1 in 30) [{"name":"Yes","probability":0.03,"type":"PROBABILITY"},{"name":"No","probability":0.97,"type":"PROBABILITY"}] 2
28 Human extinction by 2100 as a result of the single biggest natural pandemic https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates Actual estimate: 0.05% This is the median. Beard et al.'s appendix says "Note that for these predictions no time frame was given." I think that that's incorrect, based on phrasings in the original source, but I'm not certain. [{"name":"Yes","probability":0.0005,"type":"PROBABILITY"},{"name":"No","probability":0.9995,"type":"PROBABILITY"}] 2
29 AI-induced existential catastrophe Existential catastrophe from naturally arising pandemics by 2120 https://futureoflife.org/2020/04/15/an-overview-of-technical-ai-alignment-in-2018-and-2019-with-buck-shlegeris-and-rohin-shah/ https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 X-risk estimates Actual estimate: ~0.01% (~1 in 10,000) [{"name":"Yes","probability":0.5,"type":"PROBABILITY"},{"name":"No","probability":0.5,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.0001,"type":"PROBABILITY"},{"name":"No","probability":0.9999,"type":"PROBABILITY"}] 2
30 Existential risk from unaligned AI over the coming 100 years Human extinction by 2100 as a result of single biggest engineered pandemic https://forum.effectivealtruism.org/posts/2sMR7n32FSvLCoJLQ/critical-review-of-the-precipice-a-reassessment-of-the-risks https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates Actual estimate: 0.05% This was a direct response to Ord's estimate. It focuses on one pathway to x-risk from AI, not all pathways (e.g., not AI misuse or risks from competition between powerful AIs). "These estimates should not be taken very seriously. I do not believe we have enough information to make sensible quantitative estimates about these eventualities. Nevertheless, I present my estimates largely in order to illustrate the extent of my disagreement with Ord’s estimates, and to illustrate the key considerations I examine in order to arrive at an estimate." In comments on the source, Rohin Shah critiques some of the inputs to this estimate, and provides his own, substantially higher estimates. This is the median. Beard et al.'s appendix says "Note that for these predictions no time frame was given." I think that that's incorrect, based on phrasings in the original source, but I'm not certain. [{"name":"Yes","probability":0.0005,"type":"PROBABILITY"},{"name":"No","probability":0.9995,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.02,"type":"PROBABILITY"},{"name":"No","probability":0.98,"type":"PROBABILITY"}] 2
31 Existential risk from AI Annual probability of an existential catastrophe arising from a global pandemic https://youtu.be/WLXuZtWoRcE?t=1229 https://www.liebertpub.com/doi/10.1089/hs.2017.0028 X-risk estimates Actual estimate: 5-30% I put the probability that [AI/AGI] is an existential risk roughly in the 30% to 5% range, depending on how the problem is phrased." I assume he means the probability of existential catastrophe from AI/AGI, not the probability that AI/AGI poses an existential risk. Actual estimate: 0.008% to 0.0000016% (between 8 x 10-5 and 1.6 x 10-8) The fact that there's a separate estimate from the same source for biowarfare and bioterrorism suggests to me that this is meant to be an estimate of the risk from a natural pandemic only. But I'm not sure. This might also include "accidental" release of a bioengineered pathogen. [{"name":"Yes","probability":0.175,"type":"PROBABILITY"},{"name":"No","probability":0.825,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.00004,"type":"PROBABILITY"},{"name":"No","probability":0.99996,"type":"PROBABILITY"}] 2
46 Civilization collapses and does not recover, or a situation where all human life ends due to nanotechnology Global human population of zero resulting from the 150 Tg of black carbon scenario in our 2007 paper https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 http://www.overcomingbias.com/2012/11/nuclear-winter-and-human-extinction-qa-with-luke-oman.html X-risk estimates Actual estimate: 0.0100% Actual estimate: 0.001-0.01% (“in the range of 1 in 10,000 to 1 in 100,000”) I think that this is Oman’s estimate of the chance that extinction would occur if that black carbon scenario occurred, rather than an estimate that also takes into account the low probability that that black carbon scenario occurs. I.e., I think that this estimate was conditional on a particular type of nuclear war occurring. But I’m not sure about that, and the full context doesn’t make it much clearer. [{"name":"Yes","probability":0.0001,"type":"PROBABILITY"},{"name":"No","probability":0.9999,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.000055,"type":"PROBABILITY"},{"name":"No","probability":0.999945,"type":"PROBABILITY"}] 2
47 Existential catastrophe from other anthropogenic risks (which includes but is not limited to nanotechnology) by 2120 Full-scale collapse of society, perhaps due to very, very widespread famine, if there's 2 degrees of warming https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=0 https://80000hours.org/podcast/episodes/mark-lynas-climate-change-nuclear-energy/ X-risk estimates Actual estimate: ~2% (~1 in 50) See this post for some commentary: [Some thoughts on Toby Ord’s existential risk estimates](https://forum.effectivealtruism.org/posts/Z5KZ2cui8WDjyF6gJ/my-thoughts-on-toby-ord-s-existential-risk-estimates#_Unforeseen__and__other__anthropogenic_risks__Surprisingly_risky_) Arden Koehler: "...do you have a guess at what degree of warming we would need to reach for the full-scale collapse of society, perhaps due to very, very widespread famine to have say a 10% chance of happening? Mark Lynas: "Oh, I think… You want to put me on the spot. I would say it has a 30 to 40% chance of happening at three degrees, and a 60% chance of happening at four degrees, and 90% at five degrees, and 97% at six degrees. [...] Maybe 10% at two degrees. [{"name":"Yes","probability":0.02,"type":"PROBABILITY"},{"name":"No","probability":0.98,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] 2
48 Total existential risk by 2120 if we just carry on as we are, with business as usual (which Ord doesn't expect us to do) Full-scale collapse of society, perhaps due to very, very widespread famine, if there's 3 degrees of warming https://80000hours.org/podcast/episodes/toby-ord-the-precipice-existential-risk-future-humanity/#estimates-for-specific-x-risks-000810 https://80000hours.org/podcast/episodes/mark-lynas-climate-change-nuclear-energy/ X-risk estimates Actual estimate: ~33% ("about one in three") Ord: ""one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen? My best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together."" Arden Koehler replies ""Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?"" Ord replies: ""That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about."" Actual estimate: 30-40% Arden Koehler: "...do you have a guess at what degree of warming we would need to reach for the full-scale collapse of society, perhaps due to very, very widespread famine to have say a 10% chance of happening? Mark Lynas: "Oh, I think… You want to put me on the spot. I would say it has a 30 to 40% chance of happening at three degrees, and a 60% chance of happening at four degrees, and 90% at five degrees, and 97% at six degrees. [...] Maybe 10% at two degrees. [{"name":"Yes","probability":0.33,"type":"PROBABILITY"},{"name":"No","probability":0.6699999999999999,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.35,"type":"PROBABILITY"},{"name":"No","probability":0.65,"type":"PROBABILITY"}] 2
49 The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe)”, assuming Human Level Machine Intelligence will at some point exist. Full-scale collapse of society, perhaps due to very, very widespread famine, if there's 4 degrees of warming https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904 https://80000hours.org/podcast/episodes/mark-lynas-climate-change-nuclear-energy/ X-risk estimates This is the mean. According to Beard et al., the question was "4. Assume for the purpose of this question that such Human Level Machine Intelligence (HLMI) will at some point exist. How positive or negative would be overall impact on humanity, in the long run? Arden Koehler: "...do you have a guess at what degree of warming we would need to reach for the full-scale collapse of society, perhaps due to very, very widespread famine to have say a 10% chance of happening? Mark Lynas: "Oh, I think… You want to put me on the spot. I would say it has a 30 to 40% chance of happening at three degrees, and a 60% chance of happening at four degrees, and 90% at five degrees, and 97% at six degrees. [...] Maybe 10% at two degrees. [{"name":"Yes","probability":0.18,"type":"PROBABILITY"},{"name":"No","probability":0.8200000000000001,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.6,"type":"PROBABILITY"},{"name":"No","probability":0.4,"type":"PROBABILITY"}] 2
50 Chance that AI, through “adversarial optimization against humans only”, will cause existential catastrophe, conditional on there not being “additional intervention by longtermists” (or perhaps “no intervention from longtermists”) Full-scale collapse of society, perhaps due to very, very widespread famine, if there's 5 degrees of warming https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism https://80000hours.org/podcast/episodes/mark-lynas-climate-change-nuclear-energy/ X-risk estimates Actual estimate: ~10% This is my interpretation of some comments that may not have been meant to be taken very literally. I think he updated this in 2020 to ~15%, due to pessimism about discontinuous scenarios: https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism?commentId=n577gwGB3vRpwkBmj Rohin also discusses his estimates here: https://futureoflife.org/2020/04/15/an-overview-of-technical-ai-alignment-in-2018-and-2019-with-buck-shlegeris-and-rohin-shah/ Arden Koehler: "...do you have a guess at what degree of warming we would need to reach for the full-scale collapse of society, perhaps due to very, very widespread famine to have say a 10% chance of happening? Mark Lynas: "Oh, I think… You want to put me on the spot. I would say it has a 30 to 40% chance of happening at three degrees, and a 60% chance of happening at four degrees, and 90% at five degrees, and 97% at six degrees. [...] Maybe 10% at two degrees. [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.9,"type":"PROBABILITY"},{"name":"No","probability":0.09999999999999998,"type":"PROBABILITY"}] 2
51 Chance that AI, through “adversarial optimization against humans only”, will cause existential catastrophe, conditional on “discontinuous takeoff” Full-scale collapse of society, perhaps due to very, very widespread famine, if there's 6 degrees of warming https://www.lesswrong.com/posts/TdwpN484eTbPSvZkm/rohin-shah-on-reasons-for-ai-optimism https://80000hours.org/podcast/episodes/mark-lynas-climate-change-nuclear-energy/ X-risk estimates Actual estimate: ~70% (but with “way more uncertainty” than his other estimates) Arden Koehler: "...do you have a guess at what degree of warming we would need to reach for the full-scale collapse of society, perhaps due to very, very widespread famine to have say a 10% chance of happening? Mark Lynas: "Oh, I think… You want to put me on the spot. I would say it has a 30 to 40% chance of happening at three degrees, and a 60% chance of happening at four degrees, and 90% at five degrees, and 97% at six degrees. [...] Maybe 10% at two degrees. [{"name":"Yes","probability":0.7,"type":"PROBABILITY"},{"name":"No","probability":0.30000000000000004,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.97,"type":"PROBABILITY"},{"name":"No","probability":0.030000000000000027,"type":"PROBABILITY"}] 2
52 Chance that we don't manage to survive that transition [to there being something that's more intelligent than humanity], being in charge of our future. A world totalitarian government will emerge during the next one thousand years and last for a thousand years or more, conditional on genetic screening for personality traits becom[ing] cheap and accurate, but the principle of reproductive freedom prevail[ing] https://80000hours.org/podcast/episodes/toby-ord-the-precipice-existential-risk-future-humanity/#transcript https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904 X-risk estimates Actual estimate: ~20% This may have been specifically if the transition happens in the net 100 years; it's possible Ord would estimate we'd have a different chance if this transition happened at a later time. "Basically, you can look at my [estimate that the existential risk from AI in the next 100 years is] 10% as, there’s about a 50% chance that we create something that’s more intelligent than humanity this century. And then there’s only an 80% chance that we manage to survive that transition, being in charge of our future. If you put that together, you get a 10% chance that’s the time where we lost control of the future in a negative way. [For people who would disagree, a question] is why would they think that we have much higher than an 80% chance of surviving this ‘passing this baton to these other entities’, but still retaining control of our future or making sure that they build a future that is excellent, surpassingly good by our own perspective? I think that the very people who are working on trying to actually make sure that artificial intelligence would be aligned with our values are finding it extremely difficult. They’re not that hopeful about it. So it seems hard to think there’s more than 80% chance, based on what we know, to get through that. Reduced from his 5% unconditional probability [{"name":"Yes","probability":0.2,"type":"PROBABILITY"},{"name":"No","probability":0.8,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.03,"type":"PROBABILITY"},{"name":"No","probability":0.97,"type":"PROBABILITY"}] 2
71 Soft AGI takeoff At least 1 billion dead as a result of the single biggest engineered pandemic before 2100 https://reducing-suffering.org/summary-beliefs-values-big-questions/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.7,"type":"PROBABILITY"},{"name":"No","probability":0.30000000000000004,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] 2
72 By at least 10 years before human-level AGI is built, debate about AGI risk will be as mainstream as global warming is in 2015 At least 1 million dead as a result of the single biggest natural pandemic before 2100 https://reducing-suffering.org/summary-beliefs-values-big-questions/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.67,"type":"PROBABILITY"},{"name":"No","probability":0.32999999999999996,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.6,"type":"PROBABILITY"},{"name":"No","probability":0.4,"type":"PROBABILITY"}] 2
73 A government will build the first human-level AGI, assuming humans build one at all At least 1 billion dead as a result of the single biggest natural pandemic before 2100 https://reducing-suffering.org/summary-beliefs-values-big-questions/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.62,"type":"PROBABILITY"},{"name":"No","probability":0.38,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
74 A government will build the first human-level AGI, assuming humans build one at all At least 1 million dead as a result of molecular nanotech weapons before 2100 http://www.stafforini.com/blog/what_i_believe/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.6,"type":"PROBABILITY"},{"name":"No","probability":0.4,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.25,"type":"PROBABILITY"},{"name":"No","probability":0.75,"type":"PROBABILITY"}] 2
75 Human-controlled AGI in expectation would result in less suffering than uncontrolled At least 1 billion dead as a result of molecular nanotech weapons before 2100 https://reducing-suffering.org/summary-beliefs-values-big-questions/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.52,"type":"PROBABILITY"},{"name":"No","probability":0.48,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] 2
76 A design very close to CEV will be implemented in humanity's AGI, conditional on AGI being built (excluding other value-learning approaches and other machine-ethics proposals) At least 1 million dead as a result of the single biggest nanotech accident before 2100 https://reducing-suffering.org/summary-beliefs-values-big-questions/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates Actual estimate: 0.5% This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.005,"type":"PROBABILITY"},{"name":"No","probability":0.995,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.05,"type":"PROBABILITY"},{"name":"No","probability":0.95,"type":"PROBABILITY"}] 2
77 A design very close to CEV will be implemented in humanity's AGI, conditional on AGI being built (excluding other value-learning approaches and other machine-ethics proposals) At least 1 billion dead as a result of the single biggest nanotech accident before 2100 http://www.stafforini.com/blog/what_i_believe/ https://www.fhi.ox.ac.uk/reports/2008-1.pdf X-risk estimates This is the median. The report about these estimates also plots the results for each question “with individual response distributions visible” in Appendix A. [{"name":"Yes","probability":0.1,"type":"PROBABILITY"},{"name":"No","probability":0.9,"type":"PROBABILITY"}] [{"name":"Yes","probability":0.01,"type":"PROBABILITY"},{"name":"No","probability":0.99,"type":"PROBABILITY"}] 2
109
110
111
112
113
114
115
116
117
118
119
120
121
122

14
data/xrisk-questions.json
View File

@ -123,8 +123,8 @@
"title": "There be zero living humans on planet earth on January 1, 2100",
"url": "https://www.metaculus.com/questions/578/human-extinction-by-2100/",
"platform": "X-risk estimates",
"author": "Metaculus responders (~)",
"description": "Actual estimate: Median: 1%. Mean: 8%.\n\nThat median and mean is as of 3rd July 2019.",
"author": "Metaculus responders (~2021)",
"description": "Actual estimate: Median: 1%. Mean: 7%.\n\nWhile the general feeling of most people, especially now that the cold war is (mostly) over, is that the risk of human extinction is extremely small, experts have assigned a significantly higher probability to the event.\n\nIn 2008 an informal poll at the Global Catastrophic Risk Conference at the University of Oxford yielded a median probability of human extinction by 2100 of 19%. Yet, one might want to be cautious when using this result as a good estimate of the true probability of human extinction, as there may be a powerful selection effect at play. Only those who assign a high probability to human extinction are likely to attend the Global Catastrophic Risk Conference in the first place, meaning that the survey was effectively sampling opinions from one extreme tail of the opinion distribution on the subject. Indeed, the conference report itself stated that the findings should be taken 'with a grain of salt'..\n\nTherefore, it is asked: will there be zero living humans on planet earth on January 1, 2100?.\n\nFor these purposes we'll define humans as biological creatures who have as their ancestors – via a chain of live births from mothers – circa 2000 humans OR who could mate with circa 2000 humans to produce viable offspring. (So AIs, ems, genetically engineered beings of a different species brought up in artificial wombs, etc. would not count.).\n\nN.B. Even though it is obviously the case that if human extinction occurs Metaculus points won't be very valuable anymore and that it will be practically impossible to check for true human extinction (zero humans left), I would like to ask people not to let this fact influence their prediction and to predict in good faith.",
"options": [
{
"name": "Yes",
@ -300,7 +300,7 @@
"stars": 2
},
{
"title": "Extremely bad (e.g. extinction)” long-run impact on humanity from “high-level machine intelligence",
"title": "Extremely bad (e.g. extinction) long-run impact on humanity from “high-level machine intelligence",
"url": "https://arxiv.org/abs/1705.08807",
"platform": "X-risk estimates",
"author": "Survey of AI experts (~2017)",
@ -488,12 +488,12 @@
"options": [
{
"name": "Yes",
"probability": 41.5,
"probability": 0.415,
"type": "PROBABILITY"
},
{
"name": "No",
"probability": -40.5,
"probability": 0.585,
"type": "PROBABILITY"
}
],
@ -764,7 +764,7 @@
"url": "https://80000hours.org/podcast/episodes/toby-ord-the-precipice-existential-risk-future-humanity/#estimates-for-specific-x-risks-000810",
"platform": "X-risk estimates",
"author": "Toby Ord (~2020)",
"description": "Actual estimate: ~33% (\"about one in three\")\n\nOrd: \"\"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\"",
"description": "Actual estimate: ~33% (\"about one in three\")\n\nOrd: \"one in six is my best guess as to the chance [an existential catastrophe] happens [by 2120]. That’s not a business as usual estimate. Whereas I think often people are assuming that estimates like this are, if we just carry on as we are, what’s the chance that something will happen?\n\nMy best guess for that is actually about one in three this century. If we carry on mostly ignoring these risks with humanity’s escalating power during the century and some of these threats being very serious. But I think that there’s a good chance that we will rise to these challenges and do something about them. So you could think of my overall estimate as being something like Russian roulette, but my initial business as usual estimate being there’s something like two bullets in the chamber of the gun, but then we’ll probably remove one and that if we really got our act together, we could basically remove both of them. And so, in some sense, maybe the headline figure should be one in three being the difference between the business as usual risk and how much of that we could eliminate if we really got our act together.\"\"\n\nArden Koehler replies \"\"Okay. So business as usual means doing what we are approximately doing now extrapolated into the future but we don’t put much more effort into it as opposed to doing nothing at all?\"\"\n\nOrd replies: \"\"That’s right, and it turns out to be quite hard to define business as usual. That’s the reason why, for my key estimate, that I make it… In some sense, it’s difficult to define estimates where they take into account whether or not people follow the advice that you’re giving; that introduces its own challenges. But at least that’s just what a probability normally means. It means that your best guess of the chance something happens, whereas a best guess that something happens conditional upon certain trends either staying at the same level or continuing on the same trajectory or something is just quite a bit more unclear as to what you’re even talking about.\"\"",
"options": [
{
"name": "Yes",
@ -780,7 +780,7 @@
"stars": 2
},
{
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe)”, assuming Human Level Machine Intelligence will at some point exist.",
"title": "The probability that the long-run overall impact on humanity of human level machine intelligence will be Extremely bad (existential catastrophe), assuming Human Level Machine Intelligence will at some point exist.",
"url": "https://docs.google.com/spreadsheets/d/1W10B6NJjicD8O0STPiT3tNV3oFnT8YsfjmtYR8RO_RI/edit#gid=511918904",
"platform": "X-risk estimates",
"author": "Survey of experts in the AI field (~2016)",