Wiki Contributions


The Learning System

In 1841, a British parliamentary commission reported that they had found citizens who had never heard of a city called London. Meeting a group of working-class boys in the street, a commission member had asked if they knew who the Queen of England was.

“Yes, sir,” said the boys, “her name is Prince Albert.”

Of course I can't tell for sure, but I get the feeling that this is the sort of thing, that to the extent it's true, those boys were screwing with the researchers.

MikkW's Shortform

Recently I was listening to a podcast with Scott Kelly on Tim Ferris's show, and Scott said something along the lines of "if you want to see what it's like to live on Mars, try living on Antarctica – it's relatively temperate compared to where you want to go". But this misses the key reason people don't live in Antarctica - it's not the harsh climate, though it is indeed harsh; it's that it gets very little sunlight. Even during the summer, which is when the sun is present, the sunlight is much weaker than the sunlight received in more equatorial locations. Sunlight is the  fundamental ingredient of life (except for when geothermal or nuclear [in the case of humans] options are present - even fossil fuels ultimately come from sunlight), so the lack of sunlight is a big detractor for living on Antarctica.

Though, now that I think about it, Mars is much further away from the sun than the earth, which ends up with (IIRC) Mars getting around a quarter of the sunlight the earth gets. I'm not sure what percentage of equatorial sunlight the Antarctic receives, but I would not be surprised it was actually even less than a quarter, which means that equatorial regions on Mars would still get more sunlight than the Antarctic; though I am not confident about this.

MikkW's Shortform

It has increasingly come to the US public's attention that the current voting system isn't providing results that are as good as we can hope for, and groups such as Andrew Yang's Forward Party have called to adopt ranked choice voting, a family of methods where voters can indicate their order of preferences for multiple candidates, not just their favorite. However, most people, when they hear "Ranked Choice Voting", have been trained to think of Instant Runoff Voting, which is one of the worst RCV methods known, and arguably even worse than the plurality system that is commonly used in the US (for one, it's expensive, which means government would have less money to spend on other programs that direly need funding; though I won't post a conclusive takedown of the method right now; do see this link for some diagrams that really provoked some thought for me).

There is one problem with IRV worth delving into here: oftentimes there will be a candidate who will beat every single other candidate in a head-to-head race. This candidate should obviously win, but IRV will not always elect that candidate, even though ranked ballots provide enough information to identify that candidate.

Likewise, even when there isn't a single candidate who fits that description, there is usually a group who each can beat any candidate outside that group. This group is called the Top Group (sometimes in technical literature it's called the Smith Set, and when there is only one candidate in the Top Group, that candidate is called the Condorcet winner). Each candidate in the Top Group has a strong position to argue they deserve to win the election; conversely, a candidate not in the Top Group should never be elected, but IRV and the current system both sometimes will do that.

The easy way to fix this is to say that we will throw away all candidates not in the Top Group before continuing to find the final winner. However, IRV would still be expensive and non-straightforward to count, and the diagrams that I linked above will still be weird. One approach, that I state is better than either the familiar FPTP or IRV systems, is to use the ranked ballots to identify each voter's favorite candidate from the Top Group, and elect the candidate in the Top Group with the most first preferences. This will be quicker and cheaper than doing a full IRV count, and will give more predictable results than IRV, and will give much better results than the current system.

That is only one approach to holding an election which ensures the winner is always in the Top Group; I do think it's an improvement over both IRV and the current system that's easy to explain, but there are still better Ranked Choice methods that always elect candidates from the Top Group.

How Many of Me Are There?

Suicide is considered more sinful than killing outsiders because suicide always reduces the size of the Meta-Person that the suicidee belonged to.


Humans are the universal economic bottleneck

I am not convinced of the statement in the title based upon the argument presented in the text. For one, I expect that very soon, trucks will be self-driving, and even if not, there is not enough generally applicable logic or variety of specific examples to support a claim of universality.

MikkW's Shortform

Glancing at Jason Crawford's latest post "In the shadow of the great war", I was pondering about the hypothesis that people are less optimistic about the effects of technological growth – which seems like a very reasonable change in perspective considering the many negative side effects that have occurred from technological growth in the past century or so. 

This gets me thinking about how we can restore people's optimism in technological growth, and eliminating negative externalities seems vital to create an environment where people can be safe in supporting future technological growth. This is a big reason why I feel quadratic funding is such an important tool for the well-being of future humanity.

If you don't know what you expect future / counterfactual versions of you want, it will be hard to co-operate, so I recommend spending time regularly reflecting on what they might want, especially in relation to things that you have done recently. Reflect on what actions you have done recently (consider both the most trivial and the most seemingly important), and ask yourself how future and counterfactual versions of you will react to finding out that (past) you had done that. If you don't get a gut feeling that what you did was bad, test it out by trying to create and simulate a specific counterfactual version of yourself that would react in a maximally horrified way, and reflect on what factors made that version of you horrified, and reflect on how likely those or similar factors could be. You could spend ~7 - 10 mins each day doing this reflection, or ~30 mins each week, to develop a habit of thinking in this way. I'd recommend starting with the daily version, so you can really get used to it, before maybe going to the weekly version, but you can start with the weekly version if that's more convenient, and get good results from that, too.

Also remember that the way other humans will treat counterfactual versions of you will depend on their predictions of what you will do in this branch of reality, so try to act in a way, that if the people interacting with the counterfactual_you predicted or learned you would act in that way, they would be maximally willing to do what counterfactual_you wants them to do.

Xylitol's Shortform

You can surround text with two asterisks (*) on each side to bold text, at least in the Markdown editor. With the rich-text editor, you can just click on the bold button.

MikkW's Shortform

I want there to be a way of telling time that is the same no matter where you are. Of course, there's UTC, but it uses the same names as the traditional locality-dependent clocks, so it can only be used unambiguously if you explicitly state you're using UTC, or you're in a context where it's understood that times are always given in UTC (in the military, "Zulu" is a codeword indicating UTC time; I wouldn't mind if people got in the habit of saying "twenty-two thirty Zulu" to refer to times, though I do worry it might seem a little weird to non-familiar people... though maybe actually less weird than my original proposal, which follows)

There are both Western and Chinese Zodiacs, which are traditionally used to split time into 12 chunks (12 months in the Western, 12 years in the Oriental). Together, there are 24 zodiac names, enough to give each UTC hour a zodiac name. The zodiac hours are the same for all people at any given moment, regardless of location:

00:00 - Aries 01:00 - Rat 02:00 - Taurus 03:00 - Ox 04:00 - Gemini 05:00 - Tiger 06:00 - Cancer 07:00 - Rabbit 08:00 - Leo 09:00 - Dragon 10:00 - Virgo 11:00 - Snake 12:00 - Libro 13:00 - Horse 14:00 - Scorpio 15:00 - Goat 16:00 - Sagittarius 17:00 - Monkey 18:00 - Capricorn 19:00 - Rooster 20:00 - Aquarius 21:00 - Dog 22:00 - Pisces 23:00 - Pig

So you might say "Let's meet at a quarter past aquarius" to say 20:15 UTC, or 2:15 PM in Pacific Summer Time. Maybe I should just encourage people to use Zulu time more often, though.

Load More