2017-12-18 RRG Notes
- Slack is the absence of binding constraints on behavior
- Slack gives you margin for error
- Slack allows you to relax and trade
- Slack gives you optionality - you can avoid bad trades and wait for better deals
- Slack gives you room to invest for the long term
- Slack gives you the room to stand by your moral code
- Related Slackness
- Slack in project management is the amount of time a task can be put off without causing further delays
- A slacker is one who has a lazy work ethic, or otherwise does not exert maximum effort
- Out To Get You and the Attack on Slack
- The world is filled with demands on your time, money and attention
- If you accede to too many small demands, you'll run out of reserves
- Acceding to a single large demand often makes the demand expand to completely take up your reserves
- Sometimes this is okay - sometimes you have to exert maximum effort
- But okay doesn't mean it's sustainable
- Most times are ordinary, so it's fine to make an ordinary effort
- "You can afford it"
- People like to tell you that "you can afford things"
- But everything you can afford takes a bite out of your reserves
- If you accept too many demands, you quickly run out of slack
- Instead of asking whether you can afford it, ask whether it's worth it to you
- Affordability tells you what you can't have, not what you should have
- The Slackless Life of Maya Millenial
- Things are really bad when the presence of slack if your life is viewed as a defection
- The world sees slack as insufficient dedication and loyalty
- If you're not putting your all into your image, your career, or your cause, you're a "failure"
- Make sure that you have slack under normal circumstances
- Respect the slack of others
- Slack is the distance from binding constraints
- Slack can be expressed as
- Slack disappears when the spare capacity of any single resource goes to zero, regardless of how much of everything else you have
- Maintaining slack requires balancing all important resources, making sure to shore up the scarcest resources first
- Concentrate on getting the resources that you need, not the resources that are easiest to obtain
- Unfortunately, for most people the scarce resources are the ones that they're the worst at acquiring, and probably have aversive thoughts about
- However, ignored constraints don't go away - they're still binding you
- When no single resource is very scarce, you can figure out exchange rates between resources to help shore up resources that you're drawing down faster
- Example: how many dollars is an hour of free time worth?
- Remember to account for intangible resources, like goodwill, when tracking slack
- The main advantage of slack is that it gives you optionality - the ability to change your plans and explore new paths
- Be very careful of highly competitive environments - they incinerate slack
- "You waste years by not being able to waste hours." -- Amos Tversky
- Privileged people ignore marginalized people
- Social justice spaces attempt to fix this with formal rules about how privileged people should respond to marginalized people
- Formal rules
- You should listen to marginalized people
- When a marginalized person calls you out, don't argue
- Believe them, apologize, and don't do it again
- When you see others doing what you were called out for doing, call them out
- I don't even agree with the formal rules here.
- Replace "marginalized people" with "House Unamerican Activities Committee" or "Central Soviet"
- I especially vehemently disagree with the obligation to call out others for engaging in the behavior that one was called out for
- It is not possible to follow these rules literally because:
- Marginalized people are not a monolith
- Marginalized people have the same range of opinions as privileged people
- When two marginalized people tell you different things, it's logically impossible to follow both directives
- Objectifying marginalized people doesn't create justice
- Since the rules are impossible to follow, what actually ends up happening is
- One opinion is lifted up as "the opinion of marginalized people"
- "Listening to marginalized people" consists of agreeing with that opinion
- Disagreeing with that opinion is equated with "talking over marginalized people"
- This results in fights over who is the "true voice" of marginalized people
- These rules also leave marginalized people open to sabotage
- People use language of call-outs to sabotage effective leaders
- Rules about shutting up and listening to marginalized people make it very difficult to counteract lies and distortions
- Rules are also exploited by abusive people
- Abusive person convinces the victim that they're a marginalized person
- Rules about listening to marginalized people prevent victim from asserting their rights
- Abusers can send victims into depressive spirals by claiming that everything that brings the victim joy is a symbol of oppression
- Abuser may separate victim from friends or allies by spreading rumors about oppressive behavior
- Rules that say that some people should unconditionally defer to others are dangerous
- Rules lack intersectionality
- No one experiences every form of oppression
- Call-outs are often between people who are oppressed in different ways
- Rules prevent groups dedicated to one form or marginalization from allying with groups dedicated to other forms of marginalization
- Need to "start listening for real" in order to do better by each other and work through conflicts in a substantive way
- Critique of The Leaning Tower of Morality
- Leaning Tower of Morality claims that morality is a result of group selection, rather than individual selection
- Conflates evolutionary self-interest with the ordinary usage of the term
- Obscures the fact that people aren't self-interested, it's genes that are self interested
- It's entirely possible for organisms to display altruistic behavior, while still having self-interested genetics
- The evolutionary origin of altruistic behavior confuses us, because we think that they're "fake" in some way if they're the result of evolution
- But that's confusing outcome and process - it's possible for a selfish process to have altruistic outcomes
- When we talk about genetic selfishness, we're talking about the process that results in our behaviors; when we talk about ordinary selfishness, we're talking about the behaviors themselves
- Even though we use the same word for it, it's not the same concept
- The opposite also applies
- Something that is rational for your genes need not be implemented in the rational part of your mind
- It's possible for "crimes of passion" to be "genetically rational" while involving no rational thought whatsoever
- Disambiguate what the right thing is from why you're set up to believe that's the right thing
- There is a complete break between ourselves and the process that created our minds - our morality stops at our thoughts and actions
- Human morality cannot and should not encompass evolution
- A neurotic desire for approval
- The world outside of our minds is fundamentally amoral
- Why do we care that there is no "true" altruism that's universal?
- Why are we so desperate to justify everything to evolution?
- Evolution is not God, nor should it be
- The notion of selfishness and unselfishness are human constructs
- In evolutionary terms, what spreads is what spreads - that's it
- We assign moral values to mechanisms by which genes spread themselves and then get confused when evolution seems to work in ways different from our moral intuitions
- Strong view weakly held describes hedgehogs better than foxes
- Foxes are better described with "weak views strongly held"
- If this is surprising to you, it's because you're more familiar with the degenerate forms of foxes and hedgehogs
- Cactus - degenerate hedgehog - strong views, strongly held
- Hedgehog - strong views, weakly held
- Fox - weak views, strongly held
- Weasel - degenerate fox - weak views, weakly held
- True foxes and hedgehogs are rare and complex individuals
- While both foxes and hedgehogs are capable of changing their minds in meaningful ways, weasels and cacti are not
- Views and holds
- The basic distinction between foxes and hedgehogs:
- Fox: knows many things
- Hedgehog: knows one big thing
- Many things refers to weak views
- One big thing: refers to strong views
- To get a hedgehog to change their mind, you have to offer a single big idea that is more powerful than the big idea they already hold
- However, this isn't as hard as it sounds because a hedgehog's "big idea" is anchored by a relatively small number of core foundational assumptions
- Break one or two of the assumptions and the hedgehog will update their entire worldview
- To get a fox to change their mind, you need to undermine beliefs in multiple ways in multiple places
- Fox ideas are anchored by multiple instances in multiple domains
- Connected by narratives and metaphors
- To get a fox to change their mind, you have to undermine many fragmentary beliefs in many places
- While each individual belief may be easier to undermine, there are a lot more beliefs that you have to attack
- Not much coherence to exploit
- It's a lot easier to make a hedgehog change their mind wholesale - if you can undermine the right one or two beliefs, you'll effect a conversion almost overnight
- The strength of views
- Beliefs aren't held as disconnected sets of atomic propositions
- Most beliefs are organized into clusters that we refer to as "views" - each view corresponds to one or more problem domains
- If all views are connected and relatively consistent, then we can refer to the connected set of views as a "world view"
- Both foxes and hedgehogs have views, but only hedgehogs have world views
- A view is a set of interdependent beliefs
- Some beliefs are axiomatic
- When they're undermined, the view collapses
- Other beliefs are peripheral, and can be discarded without discarding the view
- A strong view is one that encompasses a large set of beliefs, and is defended with the most literal interpretation imaginable
- A weak view is one that encompasses a few critical beliefs and is defended with the most robust interpretation available
- A strong view has some advantages:
- Powerful - to the extent that the view is true or unfalsifiable, it is very useful
- Offers practical prescriptions about how to live your life
- Allows you to make detailed predictions about the future (which may or may not be correct)
- Tedious to undermine, even though it is lightly held
- Opponent must pass an ideological Turing Test in order to attack the view
- Falsification must operate within the epistemology of the view
- In some ways, your opponent has to know more about the view than you do
- Unfortunately, for most views today, we have no idea what the critical central beliefs are and what the peripheral beliefs are, even for our own views
- Changing your mind
- We change our mind easily when dealing with isolated, atomic, peripheral beliefs
- When people talk about changing minds being difficult, they're really talking about changing views
- Changing a view is like replacing a program with another program
- Changing a world-view is like replacing one operating system with another
- Strong views represent a high sunk cost
- In order to change a strong view, you need to:
- Learn new habits
- Learn new patterns of thinking
- Order matters - people learn new habits first, and then change their patterns of thinking to match their new habits
- Example: debating creationists
- Trying to get a creationist to change their mind through logical debate is pointless
- Start by getting them to use innocuous tools and habits whose effectiveness depends on evolutionary thinking
- Then, once they understand the power of evolutionary processes, they'll be ready to change their minds
- Paradoxically, this means that it's easier to change a hedgehog's mind than a fox's
- Foxes don't have as much deep expertise, so it's more difficult to get them to change their mind by altering behavior
- Strong Views, Weakly Held
- Cultivate the ability to change world-views very fast
- It is much easier to detect when one of your core beliefs has been undermined than it is to determine what the core beliefs are in the first place
- This is because hedgehogs operate on habit; don't think about the fundamental logical structure of their views day-to-day
- Therefore, and opponent can look at the logical structure of their views from the outside and see weaknesses that the person holding the view may be unaware of
- A good hedgehog will realize that their opponent is a teaching aid and will cultivate the following skills:
- Recognize when their world-views have been undermined
- When a core belief in a view has been undermined, assume that all other beliefs in that view have been undermined as well
- Build a new view, on top of new habits; treat old views and beliefs as false until proven otherwise
- Beginner hedgehogs don't even realize when their world-view has been undermined
- Intermediate hedgehogs realize that their world-views have been undermined but don't automatically doubt all of the beliefs in the worldview
- Advanced hedgehogs realize that world-views have been undermined and doubt old beliefs, but can still sometimes follow old habits even as they're building a new world-view
- Weak Views, Strongly Held
- Foxes don't have strong-views, as detailed above
- They have superficial views about a variety of fields, rather than a strong view of a single field
- However, foxes hold their views strongly, in that they accept or reject locally fundamental beliefs based upon justifications from other domains (often driven by analogy or metaphor)
- Views are anchored by many independent justifications from many different domains, rather than many details from a single domain
- If the basic challenge for the hedgehog is switching world-views and updating all of the beliefs associated with that world-views, the challenge for the fox is to switch quickly from one pattern of organization of beliefs to another
- When a belief has value in many domains, a refutation in a single domain doesn't necessarily mean the view is rejected
- The difference between foxes and hedgehogs is what they do with domain-independent truths
- Hedgehogs build totalizing world-views
- Foxes build a "refactoring mindset"
- What is a refactoring mindset?
- Heuristic and Doctrinaire Religion
- The real difference between foxes and hedgehogs is how they operate when they're outside their "home" domain
- There are no pure generalists, nor are there pure specialists
- Everyone is T shaped, but we emphasize different parts of the T
- Hedgehogs are fat-stemmed Ts; explore world in a depth-first fashion
- Start with home domain, pick the next topic, understand that topic completely, pick the next topic, etc.
- Hedgehogs have thoroughly and repeatedly read a few books, and have completely ignored others as irrelevant
- Hedgehogs rely a great deal on home-domain knowledge and prefer to apply that knowledge by applying abstraction and reasoning based on abstractions
- Instead of using ad-hoc metacognition, they emphasize building systems
- Hedgehog religions are highly doctrinaire, with strong normative rules
- Foxes are fat-bar Ts; explore world in a breadth-first fashion
- Foxes don't rely too much on home-domain expertise, instead relying on ad-hoc metacognition
- Foxes have skimmed a lot of books, and read a few completely
- Apply tools and techniques from a variety of domains to the problem until they find something that works or seems to work
- Fox-type religions have lots of locally interpretable doctrine, and rely on participants using their own meta-cognition to pick out a belief-set that works for them
- Both foxes and hedgehogs have advantages and disadvantages
- Foxes
- Foxes are great at thinking in a new domain
- Can come up with analogies relating problems in the new domain to problems or solutions in the old domain
- However, fox analogies often operate at the surface level, and aren't very useful as guides to action
- Hedgehogs
- Hedgehogs are a lot slower to get started in a new domain, because they have to come up with a systematic world-view
- However, once that systematic world-view is built, actions become automatic and habit-driven, allowing them to operate much more quickly
- So, given that hedgehogs have deep domain knowledge at home, and are able to outcompete foxes once they've built a systematic understanding of new domains, is there any advantage to being a fox?
- The Tetlock Edge
- The one edge that foxes have is prediction
- Tetlock's data shows that foxes tend to be marginally less wrong at predicting the future than hedgehogs
- Where does this advantage come from?
- Foxes eschew abstraction and prefer analogy, metaphor and narrative
- Foxes are constantly doing meta-analysis with unstructured ensembles
- Abstraction can grow into doctrine - by eschewing abstraction, foxes render themselves immune to being locked into a particular way of thinking
- The lack of abstraction prevents foxes from forming definite views of fields they haven't personally experienced
- Foxes are better able to deal with disparate and contradictory pieces of data, whereas hedgehogs are more likely to discard contradictory data as noise
- The Fox/Hedgehog Duality
- Foxes don't throw away new data once an inductive generalization has been achieved
- Hedgehogs have more rigid models and throw away data that doesn't fit their worldview
- This means that foxes slowly gain an advantage as they accumulate data and can find patterns that hedgehogs miss
- However, it takes a long time for foxes to gain this advantage and in the meantime, hedgehogs can execute faster
- Nobody can be both a fox and a hedgehog - you only have a limited amount of time and computational capacity
- The big challenge, however, is to avoid falling into the degenerate forms of foxes and hedgehogs
- Dissolute Foxes and Hidebound Hedgehogs
- Strong views, strongly held: dogma
- Cognition without meta-cognition
- Elevation of fundamental beliefs to unfalsifiable sacredness
- Hedgehogs that never change their world-view
- Weak views, weakly held: wishy-washy bullshit
- Meta-cognition without cognition
- Ephemeral engagement of ideas in purely relative terms
- Foxes that become unmoored from any kind of ground truth, and become incapable of telling truth and falsehood apart
- Bullshit detection
- Foxes are bullshit-resistant
- Agnostic to detailed truths of any domain
- If they build upon a false fact, that just causes a local revision
- Hedgehogs detect bullshit
- When encountering someone with an ostensibly opposed worldview hedgehogs have to try to see how coherent the opposed world-view is
- If the other world-view doesn't meet their standard for coherence, then hedgehogs suspect that they're dealing with an insincere opponent
- My thoughts
- I think foxes have advantages and disadvantages at different stages of the OODA loop
- Foxes have an advantage in the observe and orient stages
- Hedgehogs have an advantage in the decide and act stages
