I am writing a post on the effects of this one. If anyone is interested, I will try to finish
Metaculus currently gives a 16% chance to the claim that total deaths before 2021 will be greater than 11.6 M.
Could you please provide the JHU questions and predictions for those of us who don’t want to sign up?
I suggest the question you’ve linked has an artificially low upper bound
The question has an upper bound of 100 million deaths, not cases. I don’t think that is “artificially low”.
Maybe you are confusing Hurford’s link with this old question, which does have an artificially low upper bound and deals with cases instead of deaths.
All metaculus questions are about cases, not deaths.
Most of them are, but the one Hurford linked to is explicitly about the number of deaths: “How many people will die as a result of the 2019 novel coronavirus (2019-nCoV) before 2021?”.
I am not sure where you found the claim you cite
If you look at the bottom of the page, it says that the community predicts a ~3% chance of greater than 100 million deaths. Previously, it said 2% for the same number of deaths.
Just to be absolutely clear about what I am referring to, here is a screenshot of the relevant part of the UI.
Note: despite it being kind of neat (in my humble opinion) to develop such a scoring system, and getting mixed-to-positive feedback about it, I don’t seem to have gotten attention from EA or EA-adjacent media, journalists, podcasts, etc.
Have you tried reaching out to anyone?
The WHO has now declared a global health emergency.
The opposite trend occurred for SARS (in the same class as nCoV-2019), which originally had around a 2-5% deaths/cases rate but ended up with >10% once all cases ran their full course.
In a comment from October 2019, Ben Pace stated that there is currently no actionable policy advice the AI safety community could give to the President of the United States. I’m wondering to what extent you agree with this.
If the US President or an influential member of Congress was willing to talk one-on-one with you for a couple hours on the issue of AI safety policy, what advice would you give them?
I skimmed the post, but I couldn’t find what this is responding to. Could you provide a link for context?
The founders of PETRL include Daniel Filan, Buck Shlegeris, Jan Leike, and Mayank Daswani, all of whom were students of Marcus Hutter. Brian Tomasik coined the name.
Of these five people, four are busy doing AI safety-related research. (Filan is a PhD student involved with CHAI, Shlegeris works for MIRI, Leike works for DeepMind, and Tomasik works for FRI. OTOH, Daswani works for a cybersecurity company in Australia.)
So, my guess is that they became too busy to work on PETRL, and lost interest. It’s kind of a shame, because PETRL was (to my knowledge) the only organization focused on the ethics of AI-qua-moral patient. However, it seems pretty plausible to me that the AI safety work the PETRL founders are doing now is more effective.
In July 2017, I emailed PETRL asking them if they were still active:
Dear PETRL team,
Is PETRL still active? The last blog post on your site is from December 2015, and there is no indication of ongoing research or academic outreach projects. Have you considered continuing your interview series? I’m sure you could find interesting people to talk to.
The response I received was:
Thanks for reaching out. We’re less active than we’d like to be, but have an interview in the works. We hope to have it out in the next few weeks!
That interview was never published.
I believe kbog’s combined model is one of the best attempts yet to analyze the impacts of animal production consumption. You can view his original post announcing the model here, and his most up-to-date version of the model here.
I think that the 2018-12-05 datapoint is wrong because it came from a Quora answer which was later edited. I can’t prove this, but it seems likely because the rest of the datapoints are monotonically increasing.
Most of the other data came from FB itself (as documented in the ‘raw data’ link above), so it should be pretty solid.
Hi, as the person who personally generated the wiki dump, I can assure you that the complete content of every edit revision of every article was saved, and the data is saved in an XML format that can be trivially imported into MediaWiki. Additionally, I grabbed it after site activity had already died down, but before the wiki got taken over by spambots, so the dump should be in pretty much perfect condition.
Thanks. By the way, I updated the comment you replied to.
I generated a graph of the number of EA Forum posts per year, as well as the number of new user registrations. I extracted the data using the GraphQL API.
The raw JSON data for all posts is here. I had to split the user data into two files due to upload limits. The raw JSON data for all unbanned (but otherwise unfiltered) users is here. The JSON data for all banned users is here.
Posts by month; Posts by year
(The post count includes “meta” posts.)
2019 (so far): 433
New user registrations by month; New user registrations by year. If you include banned users, you get this monstrosity.
I divided the users into four categories in order to try to make the numbers more useful. The year entries below list the frequency values in this order.
unbanned users with non-zero posts/comments and non-negative karma.
unbanned users with non-zero posts/comments and negative karma
unbanned users with zero posts/comments
That said, the database’s indices for number of posts and number of comments don’t seem to reflect reality perfectly. Take a quick look at this table for what I mean.
Note: A previous version of this comment had different values, because I mistakenly was ignoring users with null-valued karma in the database. Now I just treat them as if they had zero karma.
2014: 336; 7; 186; 1
2015: 284; 9; 538; 1
2016: 188; 36; 540; 3
2017: 224; 70; 617; 7
2018: 366; 43; 746; 21
2019 (so far): 382; 22; 652; 1750
Regarding the Effective Altruism FB group member growth over time, I was able to piece together the following graph using archived snapshots and various other sources: https://i.imgur.com/Lejj0e1.png
The raw data (including sources for each data point) is available here. If anyone has more comprehensive data, please let me know.
Based on that, I estimate (linearly interpolate) the following member counts for January 1 of each year:
You have the actual data for 2018 and 2019. If you could share the correct counts for January 1 of those years it would be nice.