2018-03-05 RRG Notes
- Weirdness is important
- Without weirdness, we wouldn't have any social progress
- Everything we take for granted about our present society was once weird
- Six stages of policy
- Unthinkable -> radical -> acceptable -> sensible -> popular -> actual policy
- We've seen this happen with many policies
- Expansion of suffrage
- Legalizing same-sex marriage
- Some good ideas are still in the radical stage
- Effective altruism
- Many people think that donating 3% of their income is a lot, much less 10% or more
- Mitigating existenial risk
- Friendly AI
- Cryonics
- "Curing death"
- I'm not convinced that all of these policies are actually good, and I haven't written down the truly controversial ones like "open borders" or "the abolishment of genered language"
- People take weird opinions less seriously
- People are less likely to believe things that sound weird to them, even when there's overwhelming evidence for it
- Social proof matters - if fewer people believe something, other people will be less likely to believe it as well
- The halo effect is real
- We can use this knowledge to our advantage by using the halo effect in reverse
- If we're normal, we can make our weird beliefs seem more normal
- Think of weirdness as a currency that you can spend
- This leads to the following actionable principles
- Recognize you only have a few "weirdness points" to spend - if you pick one weird cause and push it, you'll have more success than trying to push every weird cause simultaneously
- Spend your weirdness points effectively - If you believe in a bunch of weird things, advocate openly for the weird thing that's going to do the most good
- Clean up and look good - if you're dressing unconventionally or sloppily, that burns a lot of weirdness points for little gain
- Advocate for more "normal" policies that are almost as good - look to see if there are any policies that are within the acceptable range of the Overton Window which could be seen as partial implementation of the "weird" policy you're thinking of advocating
- Use the "foot-in-the-door" and the "door-in-face" techniques
- Foot-in-the-door technique: start with a small ask and escalate to larger requests
- Door-in-the-face technique: start with a large ask, and negotiate down, when the other person objects
- Reconsider effective altruism's clustering of beliefs
- Right now EA is associated with donating money and donating it effectively
- Less associated with career choice, veganism, and x-risk
- We should continue this compartmentalization - leave x-risk to MIRI, etc
- Ask people to be more effective with the donations they're already making rather than asking them to donate more and be more effective
- Evaluate the above with more research
- We need more evidence about the impact of weirdness on the spread of ideas
- Literature review and market research
- Weirdness as currency is not a good way to talk about weirdness
- Most successful social movements have a fair number of weird people in them
- Objectivists
- Have all sorts of weird beliefs
- Roads should be privately owned
- Aristotle is great
- Altruism is bad
- Yet an Objectivist became chairman of the Federal Reserve
- In practice, objectivists absolutely observe an economy of weirdness. If you actually talk to objectivists, they don't talk about how Aristotle was cool. They don't talk about how altruism is bad. They talk about how the Federal Reserve is debasing the currency and how everything would be better if we went back to gold
- Feminism
- Much feminist theory has been developed by communists and socialists
- The concept of "intersectionality" means that if you want to endorse gender equality, then you also need to endorse anti-racism, anti-ableism, anti-poverty, anti-LGBT, etc. etc.
- This is a good thing? Intersectionality is great as a method of analysis, but it's an actively toxic meme for building social movements
- Ozy is confusing third-wave feminism with feminism in general
- This is why feminism has been going backwards since it adopted the intersectionality meme
- Remember, reproductive rights are now more restricted than they were in 1985
- Heck this is why Occupy Wall Street was so ineffective as well, and why Black Lives Matter only became effective when they focused on one specific thing: body cams
- If I wanted to come up with a meme to destroy incipient social movements, I would be hard pressed to find one better than intersectionality
- Pretty much every good feminist writer is a "fat hairy dyke"
- Evangelical Christianity
- Has been enormously successful despite the existence of Quiverfull
- Again, in practice, when you talk to Evangelical Christians, they don't talk about Quiverfull. They talk about how being "born again" has made a huge personal improvement in their own lives, which is just normal religious talk
- For private figures, there are two important considerations
- Weirdness is relative to your social group
- If your group dresses and acts unconventionally, then you're the weird one if you dress or act conventionally
- Okay, sure, but all this means is that your group is sufficiently isolated from mainstream society that you can get away with dressing and acting unconventionally
- This is fine for hippies and punk rock, but not fine for movements that actually want to make a dent in the world
- Concern about EA as a whole becoming perceived as a bunch of rich technolibertarian programmers
- However, EA already is a bunch of rich technolibertarian programmers
- As a result, when EA people minimize weirdness, they're going to minimize it relative to their social group
- Example: atheism - was not even mentioned in the original post on weirdness points, even though it is probably the weirdest actual belief
- Yes. I do actually understate my atheism when I interact with strangers - I say I'm Hindu, or "from a Hindu tradition", because I know full well that atheism is Weird
- Public figures should consider their role
- People being interviewed or organizing meetups may want to project normality, in order to attract the greatest variety of participants
- However, writers may have an advantage in being weird
- Writers perceived as original have more influence than writers that are perceived as going on about the same things over and over again
- Wrong! Originality gets you attention, but not influence. Why did Eliezer change tacks and start banging on the AI X-Risk drum to the point of boring everyone? You have to bore your core audience in order to get your ideas out into the world
- If you endorse conventional positions, and then endorse a weird position, your audience will think you've gone crazy, rather than taking the weird position seriously
- Wrong! People took AI X-Risk a hell of a lot more seriously when Serious People like Stephen Hawking, Bill Gates and Elon Musk started talking about it
- And the reason that Serious People like Bill Gates et. al. started talking about is because a Serious Philosopher named Nick Bostrom wrote a Serious Book about it
- You can be weird in private
- I don't think the original weirdness points article was even talking about private actions
- Not everyone is weird by choice
- A trans person can't help but wear a dress in public
- Communicating well is harder for non-neurotypical people
- Yes, which is why it's even more critical for them to minimize the expenditure of their weirdness points!
- The money analogy holds - if you have medical expenses or some other fixed cost eating a large chunk of your income, then it's especially important to watch how you're spending your money elsewhere
- Model 1: Weirdness is badness
- People don't like weird things
- The only reason to be weird is that it's hard to keep your weirdness under control
- Some characteristics are like this, but it's probably not what people have in mind when they talk about spending your weirdness points wisely
- Model 2: Weirdness is rarity is bad
- Weirdness is unusual
- Being unusual is bad
- The reason to have a weird trait is because you like the trait and you want to make it less unusual
- Here weirdness suffers from a coordination problem - if everyone had the trait, then the world would be better off, but no single person can afford the reputational cost to make the trait less weird
- Model 3: Weirdness among the cool kids is bad
- This is like the last model but it explains why you want to budget your weirdness
- What matters isn't how common a trait is overall, but how common it is among cool people
- The more weird traits you have, the less cool you are, and thus the less your vote counts
- Model 4: Weirdness is divisive
- If a trait is weird, it pleases some people while scaring off the majority
- But this isn't necessarily bad, even from a selfish perspective
- The trait might please a small group a lot, while being only mildly off-putting to the majority
- Having a few really enthusiastic followers (while being alienated from the rest of society) is often financially better than everyone being equally indifferent to you
- Causes and policy views tend to fit into this category
- This can actually make spending weirdness a good deal
- Advocating for related weird or extreme policies makes you a "true believer", and can make you more liked by the group that you're appealing to
- Model 4.1: Weirdness is divisive, the goal is spreading weird traits
- So far, we've assumed the objective is to be liked or taken seriously
- What if we change the objective to ensuring that a weird trait becomes common, regardless of whether you choose to express it
- Example: let's think about an individual that believes the following "weird" ideas
- People should care a lot about animal suffering
- People should care a lot about the far future
- Cryonics should be much more common
- Public displays of affection should be normalized
- Polyphasic sleep is something everyone should try
- In order to spread a weird trait, you have to have it or associate it with yourself
- This generally promotes having or expressing lots of weird traits
- Talking about both cryonics and the far future might reduce the number of people listening, but the people who do listen will think about two of your issues, rather than just one
- The incentive is different for narrowly directed advocacy organizations and their members - there you want to stick to the one issue
- People disliking you has particularly negative effects
- If people dislike you and the trait you are trying to spread becomes associated with you, then the trait becomes associated with a dislike
- This changes the mild dislike vs. enthusiastic support calculation
- Model 5: Weirdness is local
- What matters is what people around you find weird
- You can change the people you hang out with
- Either explicitly seek out people who share your weird beliefs or just express your weird beliefs and let the filtration happen organically
- Being weird has a fixed cost - the price of finding a social group that will tolerate the weirdness, but after the cost is paid, the weird trait is free
- Might be best to spend your weirdness points as fast as possible so that you can more quickly find the people with whom you want to surround yourself
- Model 6: Weirdness as a signal
- Weirdness often signals other things about a person
- Lack of awareness
- Lack of self-control
- However, these negative aspects can be mitigated by pointing out the weirdness and presenting extenuating circumstances
- Model 7: Weirdness is honest
- Deliberately avoiding weirdness is implict misrepresentation
- Mmmm… this sounds awfully close to "radical honesty", which is both weird and bad
- Being openly weird can mark you as an honest and "authentic" person, which is beneficial in a number of circumstances
- Having no idiosyncracies is often as weird as being extremely eccentric
- Being open about your entire cluster of beliefs makes you seem less flakey or hypocritical when you change the thing you're advocating for
- Can show how both things are tied to and underlying cause or priority
- Dishonesty is confusing and tangly - need to check all of the implications of your statements
- Easier in practice because people are usually not that great about drawing inferences in the moment
- Makes it easier to get useful feedback because people know what your true priorities are
- Also makes it easier for people to exploit you, since they know what you're interested in and where your blind-spots are
- Many of these models have some truth to them, and
- Each probably applies in varying degrees to varying parts of the real world
- Hard to tell whether people should be more weird or more normal
- You should treat weirdness differently depending on your goals
- What is the optimal allocation of weirdness from a social perspective?
- Social costs of people being judged badly
- People avoid being weird in order to be judged well
- Depends on whether people are judged relatively or absolutely
- If people are judged relatively, then a decrease in status of one person is automatically an increase in status for one or more other people, and so the global outcome is neutral
- If people are judged absolutely, then one person's status can decrease without corresponding increases in other peoples statuses, resulting in changes in the overall amount of status
- Social costs of deception
- If you actually don't want to interact with people with beliefs contrary to yours, then people hiding their beliefs in order to fit in is actively detrimental to you, because it makes that discovery process more difficult
- If, on the other hand, you're more interested in smooth social interactions, then people hiding their true beliefs may be beneficial
- Signaling race
- In some contexts "not weird" is subject to constant redefinition, so being able to be "not weird" is a sign of social awareness
- This race takes some effort from weird and non-weird people alike, which would be averted if people didn't avoid weirdness
- Neutral views
- If everyone chooses one topic on which to spend their weirdness budget, and accepts the common position on every other view, then virtually all views will be dictated by conformity
- Is this necessarily a bad thing?
- Moreover, isn't this pretty much what happens today? It's impossible for someone to be well informed enough to have weird opinions about everything
- And yet, the status quo does change
- Everyone chooses one topic, but more than one person can pick the same topic
- The status quo on almost everything will reign forever
- It seems socially optimal for people to be at least weird enough for public opinion to be shaped by thought
- Have multiple non-weird views on every issue
- Economies of scale and congestion
- Having large groups of people with the same views and tastes makes it easier for those people to consume entertainment
- On the other hand, for goods where there are not economies of scale (like parks, etc) it's good for people to have more weird tastes, so that there isn't as much congestion and competition for finite resources
- Standards
- Having weird preferences for e.g. keyboard layouts can impose a cost when others expect and accomodate only non-weird preferences
- Language barriers impose costs on people on both sides
- Variety
- Weirdness offers variety
- Some people like variety for its own sake
- Weirdness offers robustness - diversity of ideas means that when common ideas fail, there will be people with ideas and strategies that still work
- Information
- Honesty about weirdness is useful for improving policy
- Seems unclear what level of weirdness is best
- The main problem both here and with the previous piece is that everyone is treating weirdness as this one-dimensional metric
- You can't talk about weirdness in the abstract - the specific subject that you have weird views on matters a lot
- Example: using Linux on the desktop, and having white supremacist views on races are both weird
- Both consist of a tiny tiny minority
- However, the reaction of the world to your weird views will be very different
- This entire discussion about weirdness falls into the standard rationalist trap of coming up with extremely convoluted Grand Unified Theories, when it's much more tractable to think about different domains as being separate magisteria and having different standards for each
- Traditional ways of signalling
- Intelligence: complicated arguments and large vocabularies
- Health: sports achievements, heavy drink and long hours
- Wealth: expensive clothes, trips, etc
- However we have better ways of signalling now
- IQ tests for intelligence
- Medical tests for health
- Bank statements for wealth
- So why do the traditional ways of signalling persist
- Inertia
- Signaling equilibria require complex coordination and those who try to unilaterally change them seem nonconformist and clueless
- Hypocrisy
- Ancient and continuing norms against bragging push us to find plausible deniability for our brags
- We can pretend that large vocabularies convey information, that sports are just for fun and that expensive clothes are prettier or more comfortable
- It's much harder to find an excuse to talk about your IQ score or your bank statement
- Tyler Cowen argues that competency based signals are more efficient than traditional educational signals, and therefore competency based signals should win out over traditional education fairly rapidly
- But schooling isn't about education - competency-based learning divorces education from its normal social conformity context
- Employers, in practice, don't want their employees to have much initiative or independence
- Therefore, success in traditional schooling is a better indicator of workplace success than competency-based learning
