Have epistemic conditions always been this bad? 2020-01-25T04:42:52.190Z · score: 143 (49 votes)
Against Premature Abstraction of Political Issues 2019-12-18T20:19:53.909Z · score: 61 (20 votes)
What determines the balance between intelligence signaling and virtue signaling? 2019-12-09T00:11:37.662Z · score: 66 (26 votes)
Ways that China is surpassing the US 2019-11-04T09:45:53.881Z · score: 54 (22 votes)
List of resolved confusions about IDA 2019-09-30T20:03:10.506Z · score: 98 (33 votes)
Don't depend on others to ask for explanations 2019-09-18T19:12:56.145Z · score: 78 (25 votes)
Counterfactual Oracles = online supervised learning with random selection of training episodes 2019-09-10T08:29:08.143Z · score: 45 (12 votes)
AI Safety "Success Stories" 2019-09-07T02:54:15.003Z · score: 104 (31 votes)
Six AI Risk/Strategy Ideas 2019-08-27T00:40:38.672Z · score: 61 (30 votes)
Problems in AI Alignment that philosophers could potentially contribute to 2019-08-17T17:38:31.757Z · score: 84 (35 votes)
Forum participation as a research strategy 2019-07-30T18:09:48.524Z · score: 113 (38 votes)
On the purposes of decision theory research 2019-07-25T07:18:06.552Z · score: 65 (21 votes)
AGI will drastically increase economies of scale 2019-06-07T23:17:38.694Z · score: 42 (16 votes)
How to find a lost phone with dead battery, using Google Location History Takeout 2019-05-30T04:56:28.666Z · score: 52 (25 votes)
Where are people thinking and talking about global coordination for AI safety? 2019-05-22T06:24:02.425Z · score: 95 (33 votes)
"UDT2" and "against UD+ASSA" 2019-05-12T04:18:37.158Z · score: 49 (16 votes)
Disincentives for participating on LW/AF 2019-05-10T19:46:36.010Z · score: 79 (34 votes)
Strategic implications of AIs' ability to coordinate at low cost, for example by merging 2019-04-25T05:08:21.736Z · score: 57 (23 votes)
Please use real names, especially for Alignment Forum? 2019-03-29T02:54:20.812Z · score: 40 (13 votes)
The Main Sources of AI Risk? 2019-03-21T18:28:33.068Z · score: 78 (33 votes)
What's wrong with these analogies for understanding Informed Oversight and IDA? 2019-03-20T09:11:33.613Z · score: 39 (9 votes)
Three ways that "Sufficiently optimized agents appear coherent" can be false 2019-03-05T21:52:35.462Z · score: 69 (18 votes)
Why didn't Agoric Computing become popular? 2019-02-16T06:19:56.121Z · score: 52 (15 votes)
Some disjunctive reasons for urgency on AI risk 2019-02-15T20:43:17.340Z · score: 38 (11 votes)
Some Thoughts on Metaphilosophy 2019-02-10T00:28:29.482Z · score: 57 (16 votes)
The Argument from Philosophical Difficulty 2019-02-10T00:28:07.472Z · score: 49 (15 votes)
Why is so much discussion happening in private Google Docs? 2019-01-12T02:19:19.332Z · score: 87 (26 votes)
Two More Decision Theory Problems for Humans 2019-01-04T09:00:33.436Z · score: 59 (20 votes)
Two Neglected Problems in Human-AI Safety 2018-12-16T22:13:29.196Z · score: 81 (28 votes)
Three AI Safety Related Ideas 2018-12-13T21:32:25.415Z · score: 63 (25 votes)
Counterintuitive Comparative Advantage 2018-11-28T20:33:30.023Z · score: 78 (30 votes)
A general model of safety-oriented AI development 2018-06-11T21:00:02.670Z · score: 71 (24 votes)
Beyond Astronomical Waste 2018-06-07T21:04:44.630Z · score: 95 (42 votes)
Can corrigibility be learned safely? 2018-04-01T23:07:46.625Z · score: 75 (26 votes)
Multiplicity of "enlightenment" states and contemplative practices 2018-03-12T08:15:48.709Z · score: 101 (26 votes)
Online discussion is better than pre-publication peer review 2017-09-05T13:25:15.331Z · score: 18 (15 votes)
Examples of Superintelligence Risk (by Jeff Kaufman) 2017-07-15T16:03:58.336Z · score: 5 (5 votes)
Combining Prediction Technologies to Help Moderate Discussions 2016-12-08T00:19:35.854Z · score: 13 (14 votes)
[link] Baidu cheats in an AI contest in order to gain a 0.24% advantage 2015-06-06T06:39:44.990Z · score: 14 (13 votes)
Is the potential astronomical waste in our universe too small to care about? 2014-10-21T08:44:12.897Z · score: 25 (27 votes)
What is the difference between rationality and intelligence? 2014-08-13T11:19:53.062Z · score: 13 (13 votes)
Six Plausible Meta-Ethical Alternatives 2014-08-06T00:04:14.485Z · score: 50 (51 votes)
Look for the Next Tech Gold Rush? 2014-07-19T10:08:53.127Z · score: 44 (39 votes)
Outside View(s) and MIRI's FAI Endgame 2013-08-28T23:27:23.372Z · score: 16 (19 votes)
Three Approaches to "Friendliness" 2013-07-17T07:46:07.504Z · score: 20 (23 votes)
Normativity and Meta-Philosophy 2013-04-23T20:35:16.319Z · score: 12 (14 votes)
Outline of Possible Sources of Values 2013-01-18T00:14:49.866Z · score: 14 (16 votes)
How to signal curiosity? 2013-01-11T22:47:23.698Z · score: 21 (22 votes)
Morality Isn't Logical 2012-12-26T23:08:09.419Z · score: 19 (35 votes)
Beware Selective Nihilism 2012-12-20T18:53:05.496Z · score: 41 (45 votes)


Comment by wei_dai on What is the safe in-person distance for COVID-19? · 2020-03-30T03:00:20.338Z · score: 17 (6 votes) · LW · GW

Another reference (being reported in the news): Turbulent Gas Clouds and Respiratory Pathogen Emissions - Potential Implications for Reducing Transmission of COVID-19

Given various combinations of an individual patient’s physiology and environmental conditions, such as humidity and temperature, the gas cloud and its payload of pathogen-bearing droplets of all sizes can travel 23 to 27 feet (7-8 m). [...] Eventually the cloud and its droplet payload lose momentum and coherence, and the remaining droplets within the cloud evaporate, producing residues or droplet nuclei that may stay suspended in the air for hours, following airflow patterns imposed by ventilation or climate-control systems.

So indoors basically no distance is safe, outdoors maybe 10 meters is safe if people aren't up/down wind of each other.

Comment by wei_dai on Ways that China is surpassing the US · 2020-03-28T20:59:42.737Z · score: 2 (1 votes) · LW · GW

The US is losing its world superpower status due to its failure to lead on the Covid-19 crisis – and this time, it might not recover

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-27T02:10:31.389Z · score: 6 (5 votes) · LW · GW

I did sell some of the puts, but not enough of them and not near enough to the bottom to not leave regrets. I definitely underestimated how fast and strong the monetary and fiscal responses were, and paid too much attention to epidemiological discussions relative to developments on those policy fronts. (The general lesson here seems to be that governments can learn to react fast on something they have direct experience with, e.g., Asian countries with SARS, the US with the 2008 financial crisis.) I sold 1/3 of remaining puts this morning at a big loss (relative to paper profits at the market bottom) and am holding the rest since it seems like the market has priced in the policy response but is being too optimistic about the epidemiology. The main reason I sold this morning is that the Fed might just "print" as much money as needed to keep the market at its current level, no matter how bad the real economy gets.

Comment by wei_dai on What is the safe in-person distance for COVID-19? · 2020-03-27T01:21:57.898Z · score: 5 (3 votes) · LW · GW

I was thinking of thresholds that were more like “at least 12 feet apart, maybe 20 feet”, with nobody touching any objects.

My point is that since 45 out of 60 people were infected and they were spaced out, the farthest person infected in that group must have been quite far from the source of the infection, and keeping a even longer distance to be safe is probably impractical for most indoor spaces.


(Also, I’m assuming this is all outdoors)

Didn't notice this part earlier. I would be much less worried outdoors where virus particles are more likely to disperse instead of hang around, but don't have any quantitative answers to offer.

Comment by wei_dai on What is the safe in-person distance for COVID-19? · 2020-03-27T01:10:51.726Z · score: 13 (5 votes) · LW · GW

From "A choir group had 60 people show up for practice. Now 45 are sick.":

Ruth Backlund, a co-president at the Skagit Valley Chorale, said the group was monitoring public health guidelines at the time of the practice and had asked people to stay home if they showed even minor signs of illness. The group gathered in rows facing a piano and a choir director. They were all in individual chairs and had space to keep separated. Ms. Backlund had made sure there were extra soap dispensers in the bathrooms for people to wash their hands.

“Nobody was sick. Nobody touched anybody. Nobody shook hands. Nobody hugged everybody like you might do in a group. There was none of that,” Ms. Backlund said.

Given that they were spaced out and 1 asymptomatic person probably infected all 45 out of 60 in what must be a reasonably large room, it seems just impractical to keep sufficient distance to be safe indoors.

Comment by wei_dai on March Coronavirus Open Thread · 2020-03-26T17:04:35.546Z · score: 7 (3 votes) · LW · GW

Comment by wei_dai on March Coronavirus Open Thread · 2020-03-25T21:12:49.930Z · score: 15 (6 votes) · LW · GW

Does anyone have thoughts on the recent Oxford study that claims that only a very small minority of infections lead to hospitalization or death, and that >50% of the UK population is already infected?

Comment by wei_dai on Authorities and Amateurs · 2020-03-25T14:53:56.561Z · score: 10 (6 votes) · LW · GW

Right, I have some uncertainty that there's a reasonable explanation along those lines, which is (in part) why I said "I’m not suggesting we should draw strong conclusions from this." But I cringe at public officials, including public health officials doing press conferences while not observing social distancing guidelines that they themselves are promoting. Wouldn't it actually be more effective PR to show people what to do and signal that it's something they themselves are taking seriously? On a positive note on this front, I noticed that yesterday Trump and other top officials did a Fox interview / virtual townhall outdoors while being appropriately spaced out. I can't see how it would have been worse PR if all the officials had done that to begin with.

Comment by wei_dai on What should we do once infected with COVID-19? · 2020-03-25T14:47:21.823Z · score: 11 (3 votes) · LW · GW

Here's a negative study about hydroxychloroquine that just came out: (archive link because the website seems to be down)

Comment by wei_dai on Authorities and Amateurs · 2020-03-25T06:27:32.453Z · score: 10 (4 votes) · LW · GW

Neil Ferguson and Ian Lipkin, heads of their respective prominent epidemiological institutes, have caught or are suspected of having caught COVID-19. Being a top epidemiologist seems to be a strong risk factor for being infected, given that fewer than 1% of the English speaking world has probably been infected so far and there are not that many epidemiologists who are as prominent as these two. (This seems worth noting here as a surprising/interesting fact to update on. I'm not suggesting we should draw strong conclusions from this.)

ETA: See also my previous comment about a couple of authorities on economics.

Comment by wei_dai on Open & Welcome Thread - March 2020 · 2020-03-24T00:20:44.073Z · score: 6 (4 votes) · LW · GW

It looks like even Larry Summers doesn't understand comparative advantage. (Alternatively he's prioritizing virtue signaling over intelligence signaling.)

Thoughts at the end of a long week:

Why can’t the greatest economy in the history of the world produce swabs, face masks and ventilators in adequate supply?

And then there's Robert Reich, in a now deleted tweet (archived here):

Average hedge fund down 9% this year so far.

S&P 500 down 24% so far.

No way hedges could do this without inside information.

When this emergency is over, hedge funds must be investigated.

(For those who don't recognize the names, Larry Summers is "Charles W. Eliot Professor and President Emeritus at Harvard. Secretary of the Treasury for President Clinton and the Director of the NEC for President Obama" and Robert Reich is Professor of Public Policy at Berkeley and was Secretary of Labor for Clinton.)

Comment by wei_dai on March Coronavirus Open Thread · 2020-03-23T05:32:08.248Z · score: 5 (3 votes) · LW · GW

Suppose one of the drugs under investigation for COVID-19 effectively reduces its morbidity and mortality. For simplicity let's say that it quickly and completely cures 90% (or X%) of patients but does nothing to the remaining 10%, and we manage to scale up manufacturing enough to be able to treat everyone with the drug. Would the major developed countries then decide to relax the current mitigation/suppression policies and let the pandemic run its course (thereby accepting the morbidity/mortality of the refractory 10%) in order to revive their economies? What do you think the threshold for X would be for the major developed countries to do this?

Comment by wei_dai on Coronavirus: Justified Practical Advice Thread · 2020-03-22T11:47:09.707Z · score: 14 (9 votes) · LW · GW

Check the active ingredient of your hand sanitizer. You may have one that contains benzalkonium chloride, which does not work against at least one type of coronavirus. (This turns out to be the case for one of the hand sanitizers I've been using.) Get one that has "greater than 60% ethanol or 70% isopropanol" instead, as recommended by CDC.

ETA: I didn't bother to check this until today, because why would the FDA approve or fail to pull from the market, hand sanitizers that don't work against an important group of viruses? And why hasn't there been any articles in the MSM about this?

Comment by wei_dai on Matthew Barnett's Shortform · 2020-03-22T09:34:06.520Z · score: 5 (3 votes) · LW · GW

So many people have bought into the “Don’t worry about it” syndrome as a case of pretending to be wise, that I have become more pessimistic about humanity correctly responding to global catastrophic risks in the future.

See also this story which gives another view of what happened:

Most importantly, Italy looked at the example of China, Ms. Zampa said, not as a practical warning, but as a “science fiction movie that had nothing to do with us.” And when the virus exploded, Europe, she said, “looked at us the same way we looked at China.”

BTW can you say something about why you were optimistic before? There are others in this space who are relatively optimistic, like Paul Christiano and Rohin Shah (or at least they were - they haven't said whether the pandemic has caused an update), and I'd really like to understand their psychology better.

Comment by wei_dai on Coronavirus: Justified Practical Advice Thread · 2020-03-22T07:30:44.837Z · score: 2 (1 votes) · LW · GW

The second time, I meant "increase the oven temperature or time enough to thoroughly heat the food including the middle/inside of the meals to 70C for 30 minutes." So the idea is that if you're not sure only the outside of the meals may be contaminated, you need to increase the oven temperature/time but I can't tell you by how much because that depends on the specific food, so you'll have to figure that out yourself or use a meat thermometer.

Comment by wei_dai on What should we do once infected with COVID-19? · 2020-03-22T07:18:29.410Z · score: 5 (3 votes) · LW · GW

I've added some information about possible side effects to my comment. Obviously "with medical supervision" would be preferable, so sure talk to your doctor on the phone about it first if you can. (I think visiting a doctor's office is too risky at this point.) But if your doctor can't or won't talk to you about taking chloroquine/hydroxychloroquine, and you don't have preexisting conditions that make chloroquine/hydroxychloroquine more dangerous for you, it seems to me safer to take it than not. Unfortunately I'm unable to find quantitative information about the risk of side effects (UpToDate says "Frequency not defined" under "Adverse Reactions"), so it's hard to make a really informed decision about this. Perhaps to be safer, one could take chloroquine/hydroxychloroquine at home at a lower dosage than is recommended for severely sick hospitalized patients? Would you agree with that, or do you think "young and healthy" should refrain from taking any dosage, absent medical supervision? If so, what is that based on? (E.g., are you a doctor with first-hand experience or some other source of information about chloroquine side-effects?)

Comment by wei_dai on Coronavirus: Justified Practical Advice Thread · 2020-03-22T06:03:20.826Z · score: 17 (6 votes) · LW · GW

When getting restaurant takeout or delivery, put the food in an oven and bake at at least 70C/160F for 30 minutes to kill the virus. To minimize contamination, I put the whole package containing the meal into the oven, wash my hands, then operate the oven. (This assumes that only the packaging and outside of the food is likely to be contaminated, because the inside is hot enough to kill the virus already. If this is not true for the meal you're getting, increase the oven temperature or time enough to thoroughly heat the food to 70C for 30 minutes.)

This also works for mask reuse (which is where I got the idea from originally). (This article says "The new coronavirus is sensitive to heat. It can effectively inactivate the new coronavirus by heating for 30 minutes at 56 degrees Celsius." but does not cite a source for this. I wasn't able to find a study for the COVID-19 virus, but did find the following data for SARS:

Comment by wei_dai on March 20th: Daily Coronavirus Links · 2020-03-21T21:53:46.193Z · score: 9 (4 votes) · LW · GW

From the Long Now article:

The fast-and-stealthy scenario is consistent with the evidence reported to date from China, Italy, and isolated cruise ships. Consider the Diamond Princess as an extreme example of what can happen with an older population and delayed social distancing: 3,711 people on board, about 700 had the virus when tested, 400 of these were asymptomatic, 300 became sick, 7 died. But, how many of the 3,711 had recovered before they were tested? One of the problems is that the current tests for COVID-19 only tell us if someone is currently hosting the virus. The vast majority of people tested so far have been negative; is that because they never had the infection or is it because they had already successfully recovered? If 3,000 on the Diamond Princess recovered before testing, the mortality rate was 0.2%. That is high, but it is much better than the mortality rate of those we know caught the virus.

This seems very unlikely to me. For this to be true, the majority of humans would have to be able to develop antibodies and then clear the virus much faster than a substantial minority, and the virus would have to spread much faster than we currently think (i.e., have a much bigger R0 or shorter incubation time which would it place it outside the range of other respiratory viruses). Is there any inside view or outside view reasons to think this? E.g., has there ever been another virus like what the Long Now article suggests?

To distinguish between the most optimistic and pessimistic possibilities, we would need to measure how many people have developed an immunity to the virus. Fortunately, there are at least two different well-understood methods for developing such “serologic” tests, and the first such test has recently been developed for COVID-19. The first studies to use these tests are already underway in China and Singapore. We should know the results within a few months, if not weeks.

It seems not very decision-relevant for the vast majority of LWers to bring up this "optimistic possibility" at this point, given how unlikely it is, and how long we'd have to wait to distinguish it from the the "pessimistic possibility". (Also I don't like the subtext here subtly suggesting that until we know for sure, the two possibilities should be equally salient.) Am I missing anything, or is decision-relevance not the main reason for linking this?

Comment by wei_dai on Open & Welcome Thread - March 2020 · 2020-03-21T19:48:02.215Z · score: 12 (6 votes) · LW · GW

Due to the principle of comparative advantage, a country can be rich while being bad at a lot of things (i.e., having an absolute disadvantage in them). This is why richer countries can be worse (even much worse) at handling the coronavirus than some poorer countries. I suspect not realizing or understanding this may have contributed to complacency / insufficient alarm in a lot of people at the beginning of the current pandemic.

Comment by wei_dai on What should we do once infected with COVID-19? · 2020-03-18T09:33:17.532Z · score: 29 (11 votes) · LW · GW

Take chloroquine or hydroxychloroquine along with zinc ASAP (so you can avoid being hospitalized). ETA: Preferably after talking to your doctor about it. See references below for more details. If you didn't buy chloroquine/hydroxychloroquine ahead of time, beg your doctor for a prescription, or call doctors around you until you find someone willing to give you a prescription (because unlike China and South Korea, it doesn't seem to be part of the standard treatment for COVID-19 in the US).

ETA: Some relevant information from the CDC:

Based upon limited in-vitro and anecdotal data, chloroquine or hydroxychloroquine are currently recommended for treatment of hospitalized COVID-19 patients in several countries. Both chloroquine and hydroxychloroquine have known safety profiles with the main concerns being cardiotoxicity (prolonged QT syndrome) with prolonged use in patients with hepatic or renal dysfunction and immunosuppression but have been reportedly well-tolerated in COVID-19 patients.

[...] Hydroxychloroquine and azithromycin are associated with QT prolongation and caution is advised when considering these drugs in patients with chronic medical conditions (e.g. renal failure, hepatic disease) or who are receiving medications that might interact to cause arrythmias.

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-17T06:00:30.593Z · score: 2 (1 votes) · LW · GW

Thought process leading to this trade: I thought the coronavirus would cause a market decline and started selling my index funds, but only sold a small portion (that I bought recently) because I didn't want to incur capital gains taxes. A few days later it occured to me that I could buy puts as an alternative to selling, and when I started looking at puts I was surprised by how cheap they were relative to how much return they could generate if my prediction was correct, so I bought more than what was needed to hedge my long positions.

On another note, I've been wondering what lesson my kid should learn from this and other episodes. (I.e., what's the right/rational lesson?) Don't bother working hard (at least not for money), and just look out for easy ways to get rich quick?

Comment by wei_dai on Refactoring EMH – Thoughts following the latest market crash · 2020-03-17T04:44:50.636Z · score: 2 (1 votes) · LW · GW

although far less because it takes 4 days to be approved for option trading

Did you try calling them to expedite the request or figure out what caused the delay? I've applied for options trading at multiple brokerages and they all approved in 1 business day.

Comment by wei_dai on Refactoring EMH – Thoughts following the latest market crash · 2020-03-17T04:40:35.258Z · score: 7 (5 votes) · LW · GW

In a March 3rd post Dalio wrote:

Reactions to the virus (e.g., “social distancing”) will probably cause a big short-term economic decline followed by a rebound, which probably will not leave a big sustained economic impact. The fact of the matter is that history has shown that even big death tolls have been much bigger emotional affairs than sustained economic and market affairs. My look into the Spanish flu case, which I’m treating as our worst-case scenario, conveys this view; so do the other cases.

This is consistent with my earlier guess:

Global equity markets may have underestimated the economic effects of a potential COVID-19 pandemic because the only historical parallel to it is the 1918 flu pandemic (which is likely worse than COVID-19 due to a higher fatality rate) and stock markets didn’t drop that much. But maybe traders haven’t taken into account (and I just realized this) that there was war-time censorship in effect which strongly downplayed the pandemic and kept workers going to factories, which is a big disanalogy between the two cases, so markets could drop a lot more this time around. The upshot is that maybe it’s not too late to short the markets.

I also find it curious that unlike past major market drops (such as the 2008 financial crisis) no money manager has become famous from predicting it ahead of time and making money from it. The only one that comes close is Michael Burry but he is apparently just managing his own money now and it sounds like he just had a general bearish bet that wasn't specific to the coronavirus.

Comment by wei_dai on A Significant Portion of COVID-19 Transmission Is Presymptomatic · 2020-03-15T03:53:29.530Z · score: 11 (6 votes) · LW · GW

CNN - Infected people without symptoms might be driving the spread of coronavirus more than we realized

Comment by wei_dai on A Significant Portion of COVID-19 Transmission Is Presymptomatic · 2020-03-14T20:14:35.997Z · score: 11 (6 votes) · LW · GW

Additional evidence:

The earliest swabs were taken on day 1 of symptoms, with symptoms often being very mild or prodromal.


In all patients except one, throat swab RNA concentrations seemed to be already on the decline at the time of first presentation.

BTW, my speculation of how presymptomatic transmission happens is that, in addition to droplets caused by normal speaking, during that time the virus also causes very mild itching/tingling in the throat, which causes very mild coughs (throat clearings) that don't even get noticed as symptoms. (This is based on a family member having a cold recently, and me noticing such very mild symptoms in myself that I probably wouldn't have noticed if I hadn't been looking for them.)

ETA: The above linked paper is also interesting for these reasons:

Together, these data indicate active replication of SARS-CoV-2 in the throat during the first 5 days after symptoms onset.

When aligned to viral load courses, it seems there is no abrupt virus elimination at the time of seroconversion. Rather, seroconversion early in week 2 coincides with a slow but steady decline of sputum viral load. Whether certain properties such as glycosylation pattern at critical sites of the glycoprotein play a role in the attenuation of neutralizing antibody response needs further clarification. In any case, vaccine approaches targeting mainly the induction of antibody responses should aim to induce particularly strong antibody responses in order to be effective.

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-12T16:50:41.790Z · score: 8 (5 votes) · LW · GW

Up 2600% and 5200%. ETA: Now back down to 2300% and 4200%.

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-12T16:37:19.052Z · score: 5 (4 votes) · LW · GW

One explanation is that the deeper out-of-the-money put (which remains out-of-the-money) benefits from both a fall in the underlying security and an increase in VIX. The shallower out-of-the-money put (which became in-the-money as a result of the market drop) benefits from the former, but not so much from the latter. Maybe another way to explain it is that the deeper out-of-the-money put was more mispriced to begin with.

Comment by wei_dai on March Coronavirus Open Thread · 2020-03-12T00:20:02.472Z · score: 10 (7 votes) · LW · GW

That's a really interesting blog post, and it made me update (towards the idea that containment efforts in most countries will keep ramping up until containment actually succeeds). How did you come across it? I've been following Twitter, a couple of FB groups, and Reddit, and it didn't get linked by any of the posts I saw.

It feels to me now that flattening the curve is just a nice graphic without anyone checking the math, but I am confused that many informed-seeming experts are promoting the idea. Anything I’m missing?

I'm wondering this too.

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-11T20:59:56.691Z · score: 30 (9 votes) · LW · GW

An update on this trade in case anyone is interested. The position is now up 1500%. I also have another position which is up 2300% (it's a deeper out-of-the-money put, which I realized would be an even better idea after seeing a Facebook post by Danielle Fong). For proper calibration I should mention that a significant part of these returns is due to chance rather than skill:

  1. VIX (a measure of stock market volatility priced into options) was unreasonably low when I bought the puts (apparently because traders got used to central banks rescuing the stock market on every downturn), meaning the put options were underpriced in part due to that, but I didn't know this.
  2. Russia decided not to cooperate with Saudi Arabia in lowering oil production, in order to hurt the US shale oil industry. This is not something I could have reasonably predicted.
  3. I also didn't predict that the CDC would bungle their testing kits, and the FDA would delay independent testing by others so much, thus making containment nearly impossible in the US.
Comment by wei_dai on When to Reverse Quarantine and Other COVID-19 Considerations · 2020-03-10T20:08:41.862Z · score: 16 (5 votes) · LW · GW

This is a relevant study I found on quinine's antiviral activity (albeit on a different virus):

A previous study reported that the antimalarial drug chloroquine, a drug that shares a similar chemical property with quinine (both are alkaline in nature), inhibits pH-dependent stages of Flavivirus replication (Randolph et al., 1990). This is a likely inhibitory mechanism of quinine in our experimental model.

Comment by wei_dai on Ineffective Response to COVID-19 and Risk Compensation · 2020-03-10T02:46:08.595Z · score: 5 (3 votes) · LW · GW

Not almost none, but very few, yes. Aerosol transmissions are a thing, but for non airborne transmissible diseases, these seem to be rare compared to close-personal interaction transmissions, such as direct hand contact and touching other people or shared surfaces.

Just came across this story Coronavirus can travel twice as far as official ‘safe distance’ and stay in air for 30 minutes, Chinese study finds:

They said the study proves the importance of washing hands and wearing face masks in public places because the virus can linger in the air attached to fine droplet particles.

“Our advice is to wear a face mask all the way [through the bus ride],” they added.

[...] The researchers also found that none of those passengers in the two buses who wore face masks were infected.

They said it vindicated the decision to ask people to wear a face mask in public.

masks shouldn’t really be reused

I think they can if done properly.

Comment by wei_dai on Ineffective Response to COVID-19 and Risk Compensation · 2020-03-08T21:09:12.474Z · score: 12 (8 votes) · LW · GW

I still don't understand your model. Do you think that almost all COVID-19 infections occur by face-touching, and almost none by droplets directly landing on someone's face? (If so what's your evidence?) Because masks protect against both, whereas handwashing only protects against the former. It seems if we have 100% compliance with masks that would be more effective than 100% compliance of even very high frequency of handwashing (which would also be a lot more costly for people to follow). Do you disagree with this? (If so, this might be our major crux.)

ETA: Masks don't protect against rubbing eyes, but we can fix that by telling people to also wear goggles. Also it seems a lot easier to achieve 100% compliance with masks/goggles (at least when supplies are adequate) because it's visible and we can attach social stigma to people who don't comply.

Comment by wei_dai on Ineffective Response to COVID-19 and Risk Compensation · 2020-03-08T19:45:39.904Z · score: 23 (7 votes) · LW · GW

First, we have incredibly effective and vastly underutilized ways to prevent spread of COVID-19, namely handwashing and not touching your face.

Handwashing isn't "incredibly" effective. See this meta-analysis which concluded that for flu, (telling people to use) medical procedure masks combined with (tell people to do) hand hygiene achieved RR of .73 while hand hygiene alone had a (not statistically significant) RR of .86.

I'm not sure if there are studies about telling people to not touch their faces, but it's also probably not "incredibly" effective, judging from this.

Given that, if I propose an intervention like making homemade masks from fabric which reduced handwashing compliance by 1% (perhaps due to distracting people or making them think handwashing is less critical,) it would need to be astonishingly effective to be net positive.

This doesn't seem true judging from the above numbers. If you don't trust those numbers, can you give your own, even if they're just guesses? Right now I don't see what background assumption you might have that could make this statement true.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-04T19:07:43.013Z · score: 15 (5 votes) · LW · GW

In places with aggressive testing, like Diamond Princess and South Korea, you see much lower fatality rates, which suggests that lots of cases are mild.

With South Korea, I think most cases have not had enough time to progress to fatality yet. With Diamond Princess, there are 7 deaths out of 707 detected cases so far, with more than half of the cases still active. I'm not sure how you concluded from this "that lots of cases are mild". Please explain more? That page does say only 35 serious or critical cases, but I suspect this is probably because the passengers are now spread all over the world and updates on them (e.g. progressing to serious or critical) are no longer being provided (unless someone dies).

So plausibly there are 4x as many mild cases as severe cases. This gets us to like 3% fatality rate (again assuming no supplemental oxygen, which I don’t think is clear and I expect someone else to be able to make progress on forecasting if they want).

Also don't understand this part. "4x as many mild cases as severe cases" is compatible with what I assumed (10%-20% of all cases end up severe or critical) but where does 3% come from?

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-04T17:54:46.115Z · score: 17 (11 votes) · LW · GW

Yeah, I kind of wrote that in a hurry to highlight the implications of one particular update that I made (namely that if hospitals are overwhelmed the CFR will become much higher), and didn't mean to sound very confident or have it be taken as the LW consensus. (Maybe some people also upvoted it for the update rather than for the bottom line prediction?)

I do still stand by it in the sense that I think there's >50% chance that global death rate will be >2.5%. Instead of betting about it though, maybe you could try to convince me otherwise? E.g., what's the weakest part of my argument/model, or what's your prediction and how did you arrive at it?

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-03T21:13:35.363Z · score: 4 (2 votes) · LW · GW

9 people have died in the Seattle area now, which is higher than even what I was expecting, given that hospitals aren't overwhelmed yet, and not that many people should have gotten infected early enough to have progressed so far in their disease. (Average time from symptoms to death is 14 days.) Might be just an outlier with a nursing home having gotten infected early, but there are 2 deaths even outside that.

Comment by wei_dai on Open & Welcome Thread - February 2020 · 2020-03-03T19:57:12.804Z · score: 4 (2 votes) · LW · GW

It seems to be a good paper to consider, which I hadn't seen before.

This article may be an answer: it lumps together ICU, ventilation, and death into a “primary composite end point”. That seems like an OK proxy for “death without treatment”, right?

The number of people reaching "primary composite end point" would also probably increase without treatment though, so it can only serve as a lower bound. The same table gives 15.7% as "severe cases", so 6-16% seems a reasonable range, which is not too different from 5-20% I estimated earlier.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-03T19:21:52.072Z · score: 7 (4 votes) · LW · GW

I think you're right, I was just mistaken in assuming that Singapore tested everyone rather than only people with symptoms. However WHO has reported that 75% of asymptomatic cases detected in China develop symptoms later, so asymptomatic cases seemingly won't reduce the global fatality rate much.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-03T08:59:28.760Z · score: 4 (2 votes) · LW · GW

Hmm, I had noticed that dividing the number of serious or critical cases by the number of total cases gives less than 10%, but assumed that's because not all cases had enough time to progress to where they might become serious or critical yet, and the 10-20% was the authors adjusting for that. But I guess you're right that maybe they just based it on China's data.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-03T02:58:59.532Z · score: 8 (4 votes) · LW · GW

If hospitals are overwhelmed, which seems very likely at this point

Data to support this: Daegu has a population of 2.4 million, 3146 COVID-19 patients, and 2000 patients waiting for hospital beds, meaning it can treat .05% of the population at a time. Note that South Korea has the second highest number of hospital beds per capita in the world.

Comment by wei_dai on Ways that China is surpassing the US · 2020-03-03T00:31:02.221Z · score: 8 (4 votes) · LW · GW

This post seems to have aged well (in quite an unfortunate way).

Desire to leave for home growing among overseas Chinese as COVID-19 spreads - I can say from personal experience that this isn't just propaganda.

And from The Coronavirus Is Causing a Global Panic—but That’s a Good Thing:

In the United States, transmission is likely already far, far wider than the ninety publicly confirmed cases on Monday because of U.S. officials’ failure to do sufficient testing. “We haven’t found hundreds or thousands of cases because we’re not looking hard enough,” Lipsitch said. “We don’t have the testing capacity to find out what’s going on. We’ve looked largely at people who had a relation to China or high-risk areas.” That’s too low, he said, by several factors of ten. He called the initial U.S. response “utterly inadequate.” Lipsitch told me that the Chinese government’s response in Hubei Province, the epicenter of the coronavirus outbreak, was more thorough than the approach taken so far in the U.S. by the Trump Administration. In Guangdong province, Chinese health officials tested more than three hundred thousand patients in so-called fever clinics, where people who think they have a fever of any origin are seen. In contrast, the U.S. has been testing a handful of isolated cases. “Our government’s response was something like one per cent—or less—than what China did,” Lipsitch said.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-02T01:52:55.276Z · score: 6 (3 votes) · LW · GW

Trump will be blamed ... probably fairly judging from his performance so far

This Twitter thread has a great explanation of what I meant. I don't understand why prediction markets aren't reacting to this. (Maybe this particular issue with testing is too hard for voters to understand, but the "It's going to disappear. One day it's like a miracle, it will disappear" quote... Doesn't the attack ad write itself here?) Is this another failure of market efficiency, or is there some reasonable explanation?

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T22:25:29.879Z · score: 5 (2 votes) · LW · GW

But self-isolation of the entire population seems like an option that can be implemented by any country to slow the peak of an outbreak.

0.1% of Hubei's population have a confirmed infection, and its hospitals are already at the breaking point (even with national resources transferred into it). If this is the limit, then hospitals can treat at most 0.1% of the population per month, so it would take 50-100 months to treat the 5-10% who will require hospitalization, which is not a realistic amount of time for self-isolation. Even 10 months is not realistic (without totally ruining the economy) so it seems "slowing the peak" just won't help much.

Comment by wei_dai on Coronavirus: Justified Practical Advice Thread · 2020-03-01T21:31:39.231Z · score: 6 (3 votes) · LW · GW

When it's hard or impossible to avoid touching one's face or staying 2 meters away from others, wear a combination of ski mask, safety or medical goggles, and surgical mask. This would prevent touching one's face and having droplets land on one's face.

Comment by wei_dai on What are the long-term outcomes of a catastrophic pandemic? · 2020-03-01T20:13:48.895Z · score: 5 (3 votes) · LW · GW

You still haven't said why a few days ago you updated to only 5% chance of >50 million global deaths within 1 year. I'm really curious about that.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T19:20:36.490Z · score: 6 (3 votes) · LW · GW

That seems a bit misleading because current CFR estimates are largely driven by cases from Hubei province, where the hospitals are already overwhelmed. You could argue that there’s room for things to get worse, and I probably agree with that. But I don’t see how this consideration warrants such a huge update to the CFR estimates. It seems to be already factored in to quite a large degree.

My understanding is that China sent a large number of doctors from other provinces to Hubei (can't find the source now), and the outbreak was controlled via draconian means unlikely to be reproducible in most parts of the world. Hubei has a population of 58 million and only 66907 confirmed cases, so the hospitals were not nearly as overwhelmed as they could have been.

ETA: Looking at the Imperial paper you cited, the (uncorrected) estimated CFR being 18% was not the result of hospitals overwhelmed, but undercounting infections within Hubei (see page 10), which they then corrected to 1%. So I think this is consistent with the CFR estimate being based on hospitals not being overwhelmed, and it being much higher in the future when hospitals throughout the world will be overwhelmed.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T19:19:48.778Z · score: 4 (2 votes) · LW · GW

I didn’t know about virtually all Singapore cases requiring hospitalization.

That wasn't my point. I think they were hospitalized for isolation and observation/treatment, not because they required it. My understanding is that "severely ill" are the ones truly requiring hospitalization, i.e., would probably die without it.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T19:10:09.275Z · score: 23 (8 votes) · LW · GW

It seems like your arguments can be summed up as "if we slow the spread enough, hospitals won't be overwhelmed" but the US only has 924,107 beds in total, and if each case takes 4 weeks to recover ("People with more severe cases generally recover in three to six weeks.") we can treat 11 million people or 3.4% of the population over a year, but that would mean death rates from other diseases would rise a lot since those patients wouldn't have beds. Many countries do have a lot more beds per capita than the US (which surprises and confuses me) but presumably they're almost all being used already?

ETA: Actually the limiting factor probably isn't hospital beds but equipment for treating respiratory disease. For example according to this paper:

The median number of full-feature mechanical ventilators per 100,000 population for individual states is 19.7 (interquartile ratio 17.2-23.1)

This works out to be about 65,000 in the whole country which is a small fraction of what's needed to treat the number of people who will need them (.05 * .5 * 327e6 = 8,175,000) even spread out over a year or two.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T18:14:18.755Z · score: 4 (2 votes) · LW · GW

Is the 5-10% global mortality prediction conditional on COVID-19 infecting >10% of the world, or unconditional?

It's more or less unconditional at this point, since it's not clear what could stop the virus from infecting >10% of the world. If you watch the press conferences for the Seattle-area outbreak, the officials in charge are saying it's unlikely that they can contain it.

What do you think of the prospects for antivirals like remdesivir to be tested and mass-produced? How much could they lower CFR?

I think the prospects are good for successful test but I don't know about scaling up production. I've asked this myself in various places and have not gotten any answers.

Why do you think other predictions, such as those given by Metaculus 1, 2, 3 are much less pessimistic?

I don't have time to go through all those comments to find out where people gave their reasons. If you've read them, can you point to some that give the best arguments for their predictions? Then I can compare with my reasons...

Do you think shorting the market is a good idea still?

Yes, if my prediction is right and the market has only priced in a much lower death rate. I'm not as confident about this as I was in my original bet though. ETA: Mainly because of uncertainty about antivirals.

Comment by wei_dai on What will be the big-picture implications of the coronavirus, assuming it eventually infects >10% of the world? · 2020-03-01T17:55:40.161Z · score: 13 (7 votes) · LW · GW

Also, due to exponential spread, most of the 100 (or low hundreds) will have been infected recently, not enough time to have progressed to "severely ill" yet.