Forecasting Newsletter: September 2020. 2020-10-01T11:00:02.405Z · score: 40 (13 votes)
Forecasting Newsletter: August 2020. 2020-09-01T11:35:19.279Z · score: 22 (7 votes)
Forecasting Newsletter: July 2020. 2020-08-01T16:56:41.600Z · score: 31 (12 votes)
Forecasting Newsletter: June 2020 2020-07-01T09:32:57.248Z · score: 45 (18 votes)
Forecasting Newsletter: May 2020. 2020-05-31T12:35:36.863Z · score: 35 (16 votes)
Forecasting Newsletter: April 2020 2020-04-30T16:41:38.630Z · score: 54 (22 votes)
New Top EA Cause: International Supply Chain Accountability 2020-04-01T07:56:17.225Z · score: 26 (9 votes)
NunoSempere's Shortform 2020-03-22T19:58:54.830Z · score: 4 (1 votes)
Shapley Values Reloaded: Philantropic Coordination Theory & other miscellanea. 2020-03-10T17:36:54.114Z · score: 32 (13 votes)
A review of two books on survey-making 2020-03-01T19:11:13.828Z · score: 30 (16 votes)
A glowing review of two free online MIT Global Poverty courses 2020-01-15T11:40:41.519Z · score: 23 (17 votes)
[Part 1] Amplifying generalist research via forecasting – models of impact and challenges 2019-12-19T18:16:04.299Z · score: 53 (16 votes)
[Part 2] Amplifying generalist research via forecasting – results from a preliminary exploration 2019-12-19T16:36:10.564Z · score: 31 (11 votes)
Shapley values: Better than counterfactuals 2019-10-10T10:26:24.220Z · score: 88 (39 votes)
Why do social movements fail: Two concrete examples. 2019-10-04T19:56:02.028Z · score: 93 (50 votes)
EA Mental Health Survey: Results and Analysis. 2019-06-13T19:55:37.127Z · score: 46 (23 votes)


Comment by nunosempere on The Risk of Concentrating Wealth in a Single Asset · 2020-10-19T09:10:13.517Z · score: 1 (1 votes) · EA · GW

Historical anecdote:

Philippe LeBon was a French civil engineer working in the public engineering corps who became interested while at university in distillation as an industrial process for the manufacturing of materials such as tar and oil. He graduated from the engineering school in 1789, and was assigned to Angoulême. There, he investigated distillation, and became aware that the gas produced in the distillation of wood and coal could be useful for lighting, heating, and as an energy source in engines. He took out a patent for distillation processes in 1794, and continued his research, eventually designing a distillation oven known as the thermolamp. He applied for and received a patent for this invention in 1799, with an addition in 1801. He launched a marketing campaign in Paris in 1801 by printing a pamphlet and renting a house where he put on public demonstrations with his apparatus. His goal was to raise sufficient funds from investors to launch a company, but he failed to attract this sort of interest, either from the French state or from private sources. He was forced to abandon the project and return to the civil engineering corps. Although he was given a forest concession by the French government to experiment with the manufacture of tar from wood for naval use, he never succeed with the thermolamp, and died in uncertain circumstances in 1805.

Comment by nunosempere on Forecasting Newsletter: September 2020. · 2020-10-14T13:54:23.912Z · score: 2 (2 votes) · EA · GW


Comment by nunosempere on Announcing our summer 2020 ACE Movement Grants · 2020-10-13T07:49:53.269Z · score: 3 (2 votes) · EA · GW

Interesting that these are many small and distinct organizations, as opposed to a big one, or to many small ones which did the same thing in each context. I'm thinking of the ILO or various workers unions, which do seem to use the same playbook in different countries to a much greater extent.

Comment by NunoSempere on [deleted post] 2020-10-12T12:02:06.417Z

Epistemic status: Adversarial collaboration.

Have you considered that you might be falling prey to the unilateralist curse?

I've briefly looked at the evolution of past social movements, and I don't get the sense that doing this kind of thing leads to a social movement being very long lived. One of the long lived movements I've studied cultivated (though perhaps not consciously) the skill of having members on both sides of any political conflict. If I imagine EA being very long lived, which seems somewhat valuable, playing politics is not a thing I picture happening in that scenario. See also the intersection between politics and movement collapse scenarios

You lampshade this, but really, there is no attempt to quantify the impact of a donation to Theresa Greenfield. For example, what amount of money would one expect to move the election by how much, and how does that compare to donating that money to GiveWell. More trickily, if you donate x millions, would you expect a random Republican billionaire to match them?

Relatedly, I have the impression that bounded rationality and politics play really badly with each other. Consider that EA may only have the capacity to deal with so many headaches at a time, and that politics has the potential to bring too many headaches. It also seems plausible that playing politics might have a one time fixed positive gain in exchange for a longer tail of negative outcomes (rather than the opposite). See also Wei Dai's comment above.

Rules/Schelling points are important. "Don't do politics, ever", seems much more robust as a rule than "don't do politics when you think it's important". If I'm thinking of the rules I want a long-lived movement to have, the first one seems vastly superior.

Comment by nunosempere on Getting money out of politics and into charity · 2020-10-10T14:58:35.069Z · score: 1 (1 votes) · EA · GW

At the beginning, you might want to consider accepting donations to any charity (even if you give some suggestions), and then transferring the funds manually, to find out which charities people would want to donate to and to reduce the friction while you grow.

Comment by nunosempere on EA Forum Prize: Winners for June 2020 · 2020-10-08T11:29:34.791Z · score: 5 (5 votes) · EA · GW

Nice to see that these are back, and congratulations to the winners.

Comment by nunosempere on AMA: Leah Edgerton, Executive Director of Animal Charity Evaluators · 2020-10-07T21:45:16.819Z · score: 1 (1 votes) · EA · GW

Answers, for reference:

  • Will corporations stick to their animal welfare commitments?
  • When will specific animal free food technologies become cost-competitive with their traditional animal counterparts?
  • Timelines for cultured meat coming to market?
  • When will technology exist which allows the identification of the sex of a chicken before it hatches? When, if ever, will such a technology be adopted
  • When, if ever, will the global production and consumption of farmed animals stop growing? When will stop completely?
  • When will specific countries or states adopt legal protection for animals / farmed animals?
  • When will EAA organizations have a budget of more than $500 million? $1 billion?
  • Questions related to the pandemic.
  • Questions related to the budget of EAA organizations in the immediate future.
Comment by nunosempere on Some thoughts on the effectiveness of the Fraunhofer Society · 2020-09-30T21:08:20.460Z · score: 11 (5 votes) · EA · GW

Kudos for the brutal honesty.

Comment by nunosempere on Linch's Shortform · 2020-09-30T16:09:06.710Z · score: 1 (1 votes) · EA · GW

I second this.

Comment by nunosempere on How to think about an uncertain future: lessons from other sectors & mistakes of longtermist EAs · 2020-09-30T11:50:57.172Z · score: 3 (2 votes) · EA · GW

Strongly upvoted, very interesting. Here is a scenario under which you might be wrong:

Not making forecasts a. is easier, and b. doesn't expose you to accusations of being overconfident, uncalibrated and clueless. It is possible that this was the driving force for abandoning forecasts, and ideological justification came afterwards. This would happen through mechanisms like those outlined in Unconscious economics

You can see this dynamic going on really clearly in this Wall Street Article: Travel CFOs Hesitant on Forecasts as Pandemic Fogs Outlook , or in this model by Deloitte, whose authors, when contacted, refused to give a confidence interval.

It is still possible that they have uncovered some useful tools, but the assertion "the best way to maximize expected utility is to not think in terms of probabilities" just sounds really suspicious.

Comment by nunosempere on Data Analysis Involvement Opportunity (~10 hours) · 2020-09-30T11:27:04.687Z · score: 1 (1 votes) · EA · GW

The data in Excel seems corrupted.

Comment by nunosempere on EA Uni Group Forecasting Tournament! · 2020-09-24T18:45:55.108Z · score: 4 (3 votes) · EA · GW

One idea you might want to consider is to steal questions from some more established forecasting platforms, like Metaculus or Good Judgment Open. You could also join foretell, which has a really nice team interface, and compete on their questions.

Comment by nunosempere on Quantum computing timelines · 2020-09-16T14:24:59.319Z · score: 1 (1 votes) · EA · GW

Do you have a link for (Grace, 2020) ?

Comment by nunosempere on Why do social movements fail: Two concrete examples. · 2020-09-14T15:07:43.394Z · score: 1 (1 votes) · EA · GW

Many of my interests are related to General Semantics, so I'd like to understand it better.

Are you looking for any particular pointers?

Comment by nunosempere on Michael_Wiebe's Shortform · 2020-09-12T18:22:48.893Z · score: 1 (1 votes) · EA · GW

You might be interested in Existential Risk and Growth

Comment by nunosempere on How much does it cost to save a life in the mediterranean sea? · 2020-09-10T07:53:33.303Z · score: 8 (4 votes) · EA · GW

Incentives might be a problem, see here.

Comment by nunosempere on What are some low-information priors that you find practically useful for thinking about the world? · 2020-08-30T20:13:27.220Z · score: 1 (1 votes) · EA · GW

As a side-note, the maximum entropy principle would tell you to choose the maximum entropy prior given the information you have, and so if you intuit the information that the balls are likely to be produced by the same process, you'll get a different prior that if you don't have that information.

I.e., your disagreement might stem from the fact that the maximum entropy principle gives different answers conditional on different information.

I.e., you actually have information to differentiate between drawing n balls and flipping a fair coin n times.

Comment by nunosempere on When can Writing Fiction Change the World? · 2020-08-25T22:58:39.580Z · score: 20 (8 votes) · EA · GW

I have some notes about this topic, reproduced below, which might be interesting to you. As a thread to pull, you might want to talk with Miranda Dixon-Luinenburg (if she's available), who received a grant from the Long-Term fund to work on "[w]riting EA-themed fiction that addresses X-risk topics" (see the first link in the next section).

Otherwise, Uncle Tom's Cabin did have the effect of inspiring The Clansman and the subsequent film, which were also influential in the opposite direction.

Previous work

Bite sized case studies

The Works of Jules Verne

Object class: Safe

Containment procedure: None.

Description: Jules Verne wrote a series of highly entertaining novels which are still widely read, and inspired countless young people. I had a hunch that Von Braun (described as "the architect of the Moon landings") might have been inspired by Verne, and this hunch was confirmed:

Wernher von Braun, the German-born scientist who became critical to the success of the American space program, hoped the rockets he designed would enable space exploration as described by Verne and H. G. Wells... When von Braun’s Saturn rockets powered the US Mercury and Apollo programs, he gave credit to Verne’s self-fulfilling prophesies: “The science in 1865’s From the Earth to the Moon is nearly as accurate as the knowledge of the time permitted... He was read with great respect by working scientists, so carefully did he do his scientific homework.” According to von Braun’s accounting, “the debt modern astronauts owe Verne is apparent.” Source

Besides nudging Von Braun to have an interest in rockets (as opposed to, say, trains), the works of Jules Verne also inspired plenty of other pioneers (see Wikipedia: Cultural influence of Jules Verne for a full list). This is remarkable in light of other similarly popular writers, which variously wrote about:

  • A daring spy for the British crown fails to have a cooler greeting than Íñigo Montoya.
  • A cynical genius solves criminal mysteries through deductions which wouldn't work in real life.
  • A group of kids travel to a mysterious land which was previously ruled by a lion which represents God.
  • A kid goes to a magical school and fights a villain who wants to be immortal.
  • An attractive yet normal teenager rebels against a pastiche dystopia by going on TV.

Honorable mention:

  • A fellowship seeks to destroy an artifact which offers the user immense power, but which will corrupt and manipulate even the most innocent wielder.

A message to Garcia.

Object Class: Safe

Special Containment Procedures: None. In particular see the Streisand effect. Nonetheless, much like Atlas Shrugged, operatives are cautioned not to expose the wrong 14 year old. In case of exposure with acute negative effects, the speech A More Perfect Union may be prescribed, as it produces a similarly sized effect in a different ideological direction. Nonetheless, the effects of A Message to Garcia are more likely to be mildly positive.

Description: A Message to Garcia is a short publication which narrates a fictionalized historical depiction of a daring adventure by an American soldier. The reader will tend to regard the values of individual initiative and conscientiousness more favorably.

It may also cause the compulsion to share the publication with other people, to sympathize for those parties which are called libertarian in America and liberal in the rest of the world, and to view the works of Ayn Rand, Friedrich Hayek, et al. with less skepticism. The object has been widely shared, and continues to be read, since 1899. As it has reached the internet, containment is now unfeasible.

The Ones Who Walk Away from Omelas

Object Class: Safe.

Special Containment Procedures: None (see above). Operatives who are not negative utilitarians might want to promptly expose readers to Toby Ord's Why I'm Not a Negative Utilitarian.

Description: The Ones Who Walk Away from Omelas is a short story by famed writer Ursula K. Le Guin. Among utilitarians, it has the effects of making one wonder whether classical utilitarianism might not, after all, be wrong, and whether other ethical systems such as virtue ethics, or negative utilitarianism might be correct after all. Among nonutilitarian demographics, it also sparks heated debate.

The Fable of The Dragon Tyrant.

Object Class: Safe.

Special Containment Procedures: None (see above).

Description: The Fable of The Dragon Tyrant is a short story by Nick Bostrom, in which a dragon savages a land, and the king and the population learn to appease it and, despite it at first being seemingly impossible, kill it. The twist is that at the end of the story the author comments on how the Dragon Tyrant is similar to senescence, and by refraiming death as a Dragon which can be defeated, the reader may be able to overcome cached thoughts, instinctive turns of phrases

It is unclear what the impact of this specific story has been in the Zeitgeist, but I think it has the potential to be shared widely and be as culturally pervasive as the above works of fiction, and that this would be a good thing.

Pathways to impact, some speculation

Most social movements don't survive many generations. OpenPhilantropy looks like they want to donate all of Good Venture's capital, and it's unclear whether they will update in light of Trammel's paper on patient philantropists. If Effective Altruism ceases to exist in a generation or two, or becomes progressively more mediocre, literature can remain.

Given that generation from generation the political and technological landscape changes drastically, it might be difficult to leave specific and concrete advice for our intellectual descendants. They may discover new crucial considerations, or better technologies.

Thus, it is perhaps easier to leave literary breadcrumbs. For example, a sense of scientific wonder, the drive to make the world better and the disposition to do so efficiently, the expansion of one's moral circle. These ideas can be made formidable and long-lasting through literature. More ambitiously, one might try to pull off a Jules Verne and write stories about (beneficial) inventions which might be possible with the technology level of 50 years from now.

And literature is also a positive if your movement doesn't vanish into the ashes. The search for better turns of phrase, more worthy cognitive software, and metaphors which one would want to live by might contribute to our collective flourishing. And literature may also serve to stump, confuse and distract any adversaries, energize your allies, and enchanter the indecisive. Other models of impact exist.

The caveat is Sturgeon's law: that 90% of everything is crap. And of the fraction which is not crap, most is not actually good.

Comment by nunosempere on What are some low-information priors that you find practically useful for thinking about the world? · 2020-08-25T20:47:02.801Z · score: 1 (1 votes) · EA · GW

Yes, I think that this corresponds to the German tank problem after you see the first tank.

Comment by nunosempere on What are novel major insights from longtermist macrostrategy or global priorities research found since 2015? · 2020-08-14T15:58:58.431Z · score: 5 (3 votes) · EA · GW

there are a few outliers that only get widely recognized after decades, much longer than for typical insights

These sleeping beauties might happen more often the younger a field is. In particular, I don't particularly care that (perhaps lesser) insights are spread quickly once a field is producing a lot of papers.

Anyways, some other examples are Taylor polynomials (!) and various discoveries by Tsiolkovsky on the mechanics of space travel.

Comment by nunosempere on What are some low-information priors that you find practically useful for thinking about the world? · 2020-08-08T09:38:54.798Z · score: 3 (2 votes) · EA · GW

Yep, exactly right. 

Comment by nunosempere on What are some low-information priors that you find practically useful for thinking about the world? · 2020-08-07T17:04:08.835Z · score: 17 (10 votes) · EA · GW

With 50% probability, things will last twice as long as they already have.

In 1969, just after graduating from Harvard, Gott was traveling in Europe. While touring Berlin, he wondered how long the Berlin Wall would remain there. He realized that there was nothing special about his being at the Wall at that time. Thus if the time from the construction of the Wall until its removal were divided into four equal parts, there was a 50% chance that he was in one of the middle two parts. If his visit was at the beginning of this middle 50%, then the Wall would be there three times as long as it had so far; if his visit was at the end of the middle 50%, then the Wall would last 1/3 as long as it had so far. Since the Wall was 8 years old when he visited, Gott estimated that there was a 50% chance that it would last between 2.67 and 24 years. As it turned out, it was 20 more years until the Wall came down in 1989. This success of this prediction spurred Gott to write up his method for publication. (It appeared in the journal Nature in 1993.)

Source; see also Gott.

I have used this method with great success to estimate, among other things, the probability that friends will break up with their romantic partners.

I also carried out some experiments a while ago to find out what the prior probability was for me "being really sure about something", or the probability associated to "I would be highly surprised to learn if this were false." That is, for the feeling of being highly sure, how does that pan out?

On another direction, superforecasters have some meta-priors, such as "things will take longer than expected, and longer for larger organizations", or "things will stay mostly as they have."

Comment by nunosempere on What questions would you like to see forecasts on from the Metaculus community? · 2020-08-06T20:33:02.165Z · score: 4 (3 votes) · EA · GW

I'm also interested in questions around approval voting in general, and the Center for Election Science in particular.

Some stuff:

  • Conditional on less than 5 cities with >=50,000 people having implemented approval voting by Dec 31, 2022, what will the funding for the Center for Election Science be during 2023? Context: According to the CES's strategic plan converting 5 cities with >= 50,000 inhabitants is one of their main targets by 2022 (see p. 7). Conditional on them not achieving it, how will their funding look like? This can probably be operationalized with reference to IRS tax reports.
  • How many US cities with more than 50,000 people will have implemented approval voting by [date]?
  • What will CES funding look like in 2021, 2022, etc.
Comment by nunosempere on What questions would you like to see forecasts on from the Metaculus community? · 2020-08-06T20:27:44.951Z · score: 8 (3 votes) · EA · GW

A while ago, Leah Edgerton, of Animal Charity Evaluators, gave an AMA, and one of the questions I asked was What are some questions regarding EAA (effective animal advocacy) which are amenable to being forecasted?.

Her answer is in this video here. In short:

  • Will corporations stick to their animal welfare commitments?
  • When will specific animal free food technologies become cost-competitive with their traditional animal counterparts?
  • Timelines for cultured meat coming to market?
  • When will technology exist which allows the identification of the sex of a chicken before it hatches? When, if ever, will such a technology be adopted
  • When, if ever, will the global production and consumption of farmed animals stop growing? When will stop completely?
  • When will specific countries or states adopt legal protection for animals / farmed animals?
  • When will EAA organizations have a budget of more than $500 million? $1 billion?
  • Questions related to the pandemic.
  • Questions related to the budget of EAA organizations in the immediate future.

Operationalizing these questions, and finding out what the most useful things to forecast are may involve contacting ACE directly. For example, "corporations" is pretty general, so I imagine ACE has some particular ones in mind.

Comment by nunosempere on Will Three Gorges Dam Collapse And Kill Millions? · 2020-07-26T09:21:28.565Z · score: 35 (22 votes) · EA · GW

Good Judgement Open might have you covered here; see: Will China's Three Gorges Dam fail before 1 October 2020?.

Current crowd probability: 3%. (note the timeline).

Some comments I've curated from that question: Release appears to be controlled. Revising today after considering this further. The chance that the country doesn't divert water to prevent the dam from failing, even if upstream dams burst, seems very slim.

There have been heavy rains in the region which continue as of the time of writing. The dam is 181m high and the design maximum water level is 176m. Dams are designed to last hundreds of years, though climate change could mean that the original design assumptions have become outdated. There is a very slight chance of the dam "failing" within the next few months and releasing a sudden rapid uncontrolled flow downstream - just above zero. Some reports say the dam was built to hold 145 meters of water but actually that figure refers to the level at which water is released downstream in order to smooth out flood flows and maintain capacity in the reservoir. Discharge in recent days and weeks has been between 20 and 30 thousand cumecs, but this has gone up to 40k in the past, so there is still some cushion. Probably the greatest risk is of failure of one or more major dams upstream, unleashing a flood surge that could overtop the dam.
(Says a civil engineer)

This is a gravity dam, and it relies upon the construction itself to stand. It was 50 ft Above flood level. I see this as a concern without being a high probability event.

This question talks about the failure of a $ 32 Billion project completed in 2012. We have roughly 9 weeks from today till when the question is resolved. 5% for Yes is a good baseline to start. While the dam is currently holding more water than it is designed for, water can always be released if things get bad. The reason why they would be holding more water is in order to prevent the catchment areas from getting flooded.

This looks unlikely, but how unlikely seems difficult to estimate: on the one hand, quality of construction in China is poor and cutting corners is a way of life. On the other this is a flagship project, which means that there must have been stringent quality controls (in contrast with the standard situation in China). Unfortunately this is inside view. I did not try to make a historic review of dams failing around the world or in China. However the Banqiao dam failure in 1975 readily comes to mind: and catastrophic floods have been a common recurrence along Chinese history.

There were 3,523 incidents of dam failure from 1954 to 2013 (He et al. 2008; Zhao 2014) that caused significant loss of life and economic losses in china. This averages out to 67 dam failures, of various sizes per year. There are approx 87,000 dams in total. Given this, the three gorges dam broadly speaking has .08 chance of failing this year. There is significant flooding atm, which could increase probability of failure, but the concrete dam wall is 181 m (594 ft) high above the rock basis and has a max capacity water level of 175 m (wiki) Water level currently seems to be at 145-7 m from the articles I can find, which is well within capacity. The dam has passed quality checks, is a relatively new project (old dams fail more often) and there's a large amount of research done on seismic activity in the area.; There are rumors of buckling and deformation and a google earth image going around ( but I looked myself, and google earth currently shows shows no buckling, nor could I find any inconsistencies, so Ima say probs not, image seems fake. Long story short, is the dam gonna prevent flooding downstream? Maybe not, its effectiveness at doing so seems questionable based off the articles. This, however, isn't the issue at hand. We're asking is the worlds largest dam gonna fail in the next three months after passing safety checks, having research available about seismic activity in the area and currently within capacity? Highly unlikely. If upstream dams start to fail and/or if water levels breach capacity, then it gets more likely. But til then, low low chance.

One of the things I've been thinking of....dam failure means any amount of water that they didn't intend to let through passing the dam. So Im wondering, is there a higher percent chance of something small happening (whups, a couple gallons seeped through, or we lost a couple thousand gallons over the edge, our bad) or is it an all or nothing deal where when she goes, she goes, rip wuhan. Good point on the time it takes for flood level to get there. Also, thinking of ways it could potentially fail, I could conceivably imagine a scenario where the dam is at or slightly over capacity due to flooding, seismic activity happens thats unprecedented and the concrete slips off the bottom rock. There's history of other large concrete dams doing so ( but seems to be only at filling? There's also this article: which is recent and does mention some deformation to 'non structural parts of the dam' I don't exactly know what that means lol. And then finally, there was one scientist dude who has been talking about failure for a while, but stating cracks in the concrete during early stages of the building process and instances of substandard concrete, not buckling as the internet seems intent on portraying lol. Ah, one more thing: failures are most likely,the top of a dam.&text=National statistics show that overtopping,of all U.S. dam failures. The way dams are most likely to fail (this is US based, but i read a scientific study about chinese dams that was saying the same things) overtopping number one reason of failure, within the very low percentage chance of a dam failing. Slipping second most likely. So if the dam does fail, it's most likely gonna either overfill or slip. Will be good to pay attention to water levels upstream, precipitation, and how much output their letting through in anticipation.

This is a topic I have some subject-knowledge on, and I think the question requires clarification: The Forestry reference to "sudden, rapid, and uncontrolled release of impounded water" is included in their definition of “Dam failure” and the key element of the failure is the release of the impounded water. The dam is designed to manage a 1:100 year flood, derived statistically, by having the impoundment reserve capacity and controlled discharge of this amount of water. Floods greater than the 1:100 year value are managed by the sluice gates, turbine channel flow and, ultimately, by the dam’s spillway. The spillway is the lowest part of the dam crest and is designed to permit much larger flows (beyond the “Probable Maximum Flood”) As defined, I think the question asks whether there will be a failure of the dam which releases the water impounded below the spillway level, e.g., structural/geotechnical failure, undermining or uncontrolled bypass, which is highly unlikely. However, the question may be interpreted to ask whether the dam will be ‘overtopped’, with uncontrolled, rapid release over the spillway – which is quite probable this year. BTW the dam was built to reduce the frequency of flooding downstream, where millions have died from flooding of the Yangtze River. From a flood risk management perspective the dam is small at 1:100 year capacity. Negative press in competing or developed countries focussed on displacing 1.3M people in the interest of power production, not on flood risk management. Also, “The Interpreter” article is accurate in describing older dams in China (and around the world) as being potential ‘black swans’: these dams were often not designed to spill “probable maximum floods” and their failures may well jeopardize life downstream.

The dam is already controversial, so any story on it will be far reaching and potentially exaggerated. The Chinese have admitted to some movement to the dam but say it's within normal parameters. While under scrutiny for COVID and struggling with it's international image I'd like to think that evacuations would be in place if the risk was high. Because this may be a naive thought and catastrophic accidents have occurred in the past due to bureaucratic failings in similar regimes (think Chernobyl), I have input 2%.

The 3GD is a gravity dam, but the blocks are resting on the riverbed, not dug in. This is causing deformation throughout the structure. Construction is likely shoddy and the quality team that inspected the dam were from the same company that built it - not independently done. The CCP came out yesterday(?) to say that there is some deformation in the dam, but nothing to worry about. This alone from the CCP is unusual. In my estimation, the CCP would rather flood Wuhan downstream than see their flagship fall. That said, they may not have a choice. As the heavy rains in the region continue, water is making its way back into the Yangtze and thus the reservoir. Not all of the rain we've seen in the past has yet entered the reservoir. The water level is already at 164m, 175m is the warning of collapse, 185m and it's gone. The final and most likely catastrophic failure is the spillways being damaged - we're seeing this on at least two of the spillways on the livestream. A large chunk appears to be broken off spillway #6 and there is evidence of cavitation in spillway #2. If the damage to the spillways continues, the dam will fail, and badly.

(This last comment is from a new and unexperienced forecaster, with a Brier score worse than the aggregate. He still only gives 15%)

My own impression is that the aggregate seems correct, i.e., a 1:30 bet seems roughly fair.

I also somewhat disagree with "It is worth examining even if the risk is small;" it seems to me that decisions will be taken by the CCP, and that there is probably no leverage to be found here.

Comment by nunosempere on Forecasting Newsletter: June 2020 · 2020-07-01T13:51:42.514Z · score: 1 (1 votes) · EA · GW


Comment by nunosempere on I'm Linch Zhang, an amateur COVID-19 forecaster and generalist EA. AMA · 2020-06-30T20:54:42.966Z · score: 18 (12 votes) · EA · GW
  • If you look at your forecasting mistakes, do they have a common thread?
  • How is your experience acquiring expertise at forecasting similar/different to acquiring expertise in other domains, e.g. obscure board-games? How so?
  • Any forecasting ressources you recommend?
  • Who do you look up to?
  • How does the distribution skill / hours of effort look for forecasting for you?
  • Do you want to wax poetically or ramble disorganizedly about any aspects of forecasting?
  • Any secrets of reality you've discovered & which you'd like to share?
Comment by nunosempere on I'm Linch Zhang, an amateur COVID-19 forecaster and generalist EA. AMA · 2020-06-30T20:33:22.118Z · score: 5 (5 votes) · EA · GW

Hi Linch! So what's up with the Utilitarian Memes page? Can you tell more about it? Any deep lessons from utilitarian memes?

Comment by nunosempere on Problem areas beyond 80,000 Hours' current priorities · 2020-06-24T17:35:16.724Z · score: 6 (3 votes) · EA · GW

India v. China conflict is perhaps more immediately worrying than US v. China.

Comment by nunosempere on 2019 - Year in Review · 2020-06-08T23:21:46.907Z · score: 5 (4 votes) · EA · GW

Have you considered moving into Austria as well?

Comment by nunosempere on Forecasting Newsletter: May 2020. · 2020-06-05T21:26:26.048Z · score: 1 (1 votes) · EA · GW

Thanks. Cool username!

Comment by nunosempere on Forum update: Tags are live! Go use them! · 2020-06-02T16:29:16.390Z · score: 1 (1 votes) · EA · GW

I, too, would be curious about how to create a sequence.

Comment by nunosempere on Call notes with Johns Hopkins CHS · 2020-05-21T08:15:34.477Z · score: 4 (3 votes) · EA · GW

Thanks, this is useful.

Comment by nunosempere on Food Crisis - Cascading Events from COVID-19 & Locusts · 2020-05-16T09:08:02.119Z · score: 1 (3 votes) · EA · GW

A post by the Brookings Institution on this topic.

Comment by nunosempere on A glowing review of two free online MIT Global Poverty courses · 2020-05-15T11:48:19.729Z · score: 1 (1 votes) · EA · GW

On the clarifications: The tests are automatically graded. Also, I think I meant two courses simultaneously.

Comment by nunosempere on A glowing review of two free online MIT Global Poverty courses · 2020-05-06T21:17:01.979Z · score: 3 (2 votes) · EA · GW
  1. Yes. You could also try to go through an archived course, like this one, but I don't know whether you'll be able to access it.
  2. Yes.
  3. One could do 12-14h/week for 2 courses over a semester, but one could also do 12-14h for a single course. I would highly surprise me if one were able to take 5 courses at the same time with only 14h per week. This in general depends on how much you are willing, able and glad to do? Like, anchor on work similar to a university course?
  4. Usually two; yes.
Comment by nunosempere on Food Crisis - Cascading Events from COVID-19 & Locusts · 2020-05-03T06:59:58.468Z · score: 4 (3 votes) · EA · GW

Would you mind sharing why you thought this way back then...

Because the question asks about a very specifical technical definition of famine, and I think that the crowd forecasters were anchoring on "things will get bad", rather than on "this specific technical definition will be met". I'd appreciate it having more bins. I also looked into the FEWS reports, and none of them forecasted the highest level. Looking into the Our World in Data page on famines, the base rate isn't high.

Comment by nunosempere on Forecasting Newsletter: April 2020 · 2020-05-01T19:00:46.479Z · score: 1 (1 votes) · EA · GW

Edited again. If you want, throw me a bone: what's the last explicit probabilistic prediction you've made? Also, I liked your review on How to Measure Anything, which feels relevant to the topic at hand. NNTR.

Comment by nunosempere on Forecasting Newsletter: April 2020 · 2020-05-01T10:58:30.776Z · score: 1 (1 votes) · EA · GW

Thanks for the correction; edited.

Comment by nunosempere on Forecasting Newsletter: April 2020 · 2020-05-01T10:31:59.558Z · score: 1 (1 votes) · EA · GW

Oh hey, I've seen you around on GJOPen. Thanks for the correction; edited.

Comment by nunosempere on Food Crisis - Cascading Events from COVID-19 & Locusts · 2020-04-30T10:24:45.063Z · score: 6 (4 votes) · EA · GW

I've also looked into this, see here for some quick thoughts, and I broadly agree with your situation report.

Two useful links:

Comment by nunosempere on COVID Project idea: Transcription, translation, content reformatting, and summarization · 2020-04-27T11:37:43.765Z · score: 1 (1 votes) · EA · GW

For a concrete example of this, the Heinsberg study Wikipedia page is only available in German (and the English translation is a stub), even though it would be of broad interest to English audiences as well.

Comment by nunosempere on What will 80,000 Hours provide (and not provide) within the effective altruism community? · 2020-04-25T08:30:24.773Z · score: 3 (3 votes) · EA · GW

Yep. You can also see some other hashes here: (but they're just SHA-512/SHA-3 hashes)

Comment by nunosempere on What will 80,000 Hours provide (and not provide) within the effective altruism community? · 2020-04-24T18:40:53.083Z · score: 21 (16 votes) · EA · GW

You could consider changing the name of your organization. See also: EA Foundation/Foundational Research Institute are now the Center on Long-Term Risk (CLR).


Comment by nunosempere on Why I'm Not Vegan · 2020-04-09T18:10:56.922Z · score: 1 (8 votes) · EA · GW

So I think that once you accept a particular framing or ontology, or cluster of beliefs, vegetarianism starts to begin souding pretty obvious. One such cluster might be:

  • Moral realism: There is an objective and scientific answer to how much a pig's life is worth compared to a human. Ethics is at its best an investigation into the nature of reality, from which moral obligations follow.
  • Kant is cool. The answer to "why should I do good?" is "because I must".
  • Peter Singer ideas: Pain and suffering are extremely important. Negative utilitarianism. Sentience over sapience. Speciesim as being wrong.
  • Realizing that, deep down, care about animals a great amount.
  • ...

And you seem to be arguing from a framing similar to the above. However, that framing is not obvious, and one could adopt some other cluster of beliefs, such as:

  • Moral relativism: There isn't an objective and scientific answer to many moral questions. Many ethical questions or concepts are not well defined, and are best resolved by introspecting on your preferences. Morality is at its best is a coordination game played in good faith.
  • Gendlin is cool. The answer to "why do I strive to do good?" is "because I want", or "because I choose to".
  • Enlightenment humanism: Human flourishing. Sapience over sentience. Preference utilitarianism among humans.
  • Realizing that, deep down, you care about animals a small amount.
  • ...

And when arguing with someone which has beliefs near the second cluster, I don't think that assuming that beliefs in the first cluster are obviously right is a great tactical move (I'm ignoring audience effects). In fact, when I used to not be vegetarian, I found that kind of move to be extremely annoying, and to some extent I still do ("that guy is saying that things which took me years to understand and/or come to share, and which in some cases are still not clear to me, are obviously true?").

Instead, may I suggest a moral trade as a tactical move? (see: Morality at its best is a coordination game played in good faith)

  • You (@abrahamrowe) donate $4.3 (a factor of x10 because of your deep magnanimity) to @Jeff_Kaufman's best human existential risk reduction charity (easily another factor of x10 according to long-termist assumptions)
  • Jeff_Kaufman tries being vegetarian for a year (or changes his numbers above).

Considering this type of moral trade is possible because the original poster quantified his preferences to the best of his ability. This should be highly lauded, and gets a strong upvote from me.

Comment by nunosempere on NunoSempere's Shortform · 2020-04-09T11:45:56.822Z · score: 18 (6 votes) · EA · GW

What happened in forecasting in March 2020

Epistemic status: Experiment. Somewhat parochial.

Prediction platforms.

  • Foretold has two communities on Active Coronavirus Infections and general questions on COVID.
  • Metaculus brings us the The Li Wenliang prize series for forecasting the COVID-19 outbreak, as well as the Lockdown series and many other pandemic questions
  • PredictIt: The odds of Trump winning the 2020 elections remain at a pretty constant 50%, oscillating between 45% and 57%.
  • The Good Judgment Project has a selection of interesting questions, which aren't available unless one is a participant. A sample below (crowd forecast in parenthesis):
    • Will the UN declare that a famine exists in any part of Ethiopia, Kenya, Somalia, Tanzania, or Uganda in 2020? (60%)
    • In its January 2021 World Economic Outlook report, by how much will the International Monetary Fund (IMF) estimate the global economy grew in 2020? (Less than 1.5%: 94%, Between 1.5% and 2.0%, inclusive: 4%)
    • Before 1 July 2020, will SpaceX launch its first crewed mission into orbit? (22%)
    • Before 1 January 2021, will the Council of the European Union request the consent of the European Parliament to conclude a European Union-United Kingdom trade agreement? (25%)
    • Will Benjamin Netanyahu cease to be the prime minister of Israel before 1 January 2021? (50%)
    • Before 1 January 2021, will there be a lethal confrontation between the national military or law enforcement forces of Iran and Saudi Arabia either in Iran or at sea? (20%)
    • Before 1 January 2021, will a United States Supreme Court seat be vacated? (No: 55%, Yes, and a replacement Justice will be confirmed by the Senate before 1 January 2021: 25%, Yes, but no replacement Justice will be confirmed by the Senate before 1 January 2021: 20%)
    • Will the United States experience at least one quarter of negative real GDP growth in 2020? (75%)
    • Who will win the 2020 United States presidential election? (The Republican Party nominee: 50%, The Democratic Party nominee: 50%, Another candidate: 0%)
    • Before 1 January 2021, will there be a lethal confrontation between the national military forces of Iran and the United States either in Iran or at sea? (20%)
    • Will Nicolas Maduro cease to be president of Venezuela before 1 June 2020? (10%)
    • When will the Transportation Security Administration (TSA) next screen two million or more travelers in a single day? (Not before 1 September 2020: 66%, Between 1 August 2020 and 31 August 2020: 17%, Between 1 July 2020 and 31 July 2020: 11%, Between 1 June 2020 and 30 June 2020: 4%, Before 1 June 2020: 2%)


  • The Brookings institution, on Forecasting energy futures amid the coronavirus outbreak
  • The European Statistical Service is "a partnership between Eurostat and national statistical institutes or other national authorities in each European Union (EU) Member State responsible for developing, producing and disseminating European statistics". In this time of need, the ESS brings us inane information, like "consumer prices increased by 0.1% in March in Switzerland".
  • Famine: The famine early warning system gives emergency and crisis warnings for East Africa.
  • COVID: Everyone and their mother have been trying to predict the future of COVID. One such initiative is Epidemic forecasting, which uses inputs from the above mentioned prediction platforms.
  • On LessWrong, Assessing Kurzweil's 1999 predictions for 2019; I expect an accuracy of between 30% and 40%, based on my own investigations but find the idea of crowdsourcing the assessment rather interesting.
Comment by nunosempere on (How) Could an AI become an independent economic agent? · 2020-04-05T15:27:09.394Z · score: 2 (2 votes) · EA · GW

An example of money which nobody owns might be a bounty which nobody has claimed yet. A good example of that might be the SHA-1 collision bitcoin bounty, which could be (anonymously) claimed by anyone who could produce a SHA-1 collision.

On a larger scale, solving the Millenium Prize Problems would also give you access to a $1 million prize.

Comment by nunosempere on New Top EA Causes for 2020? · 2020-04-01T18:06:53.176Z · score: 3 (3 votes) · EA · GW

That is evil, I like it.

Comment by nunosempere on EA Survey 2019 Series: Donation Data · 2020-04-01T11:14:50.523Z · score: 2 (2 votes) · EA · GW

The SlateStarCodex survey data is useful to answer a limited form of that question (a comparison with other SSC-survey answerers).

Code here, in R, may be useful.

Comment by nunosempere on New Top EA Causes for 2020? · 2020-04-01T08:22:50.602Z · score: 6 (4 votes) · EA · GW

International Supply Chain Accountability.