2017-10-16 RRG Notes
- Air Florida Flight 90 crashed into a bridge on takeoff from Washington National Airport, killing 78
- Pilots failed to turn on engine anti-icing systems, causing loss of thrust
- Pilots positioned plane behind another jet in a mistaken attempt to de-ice that may have made the icing problem worse
- Engine pressure ratio indicator showed normal thrust, contradicting readings from other systems
- Co-pilot pointed out discrepancy, but pilot proceeded with takeoff anyway, leading to crash
- The accident was the result of a pattern of self-deception on the part of the captain, and insufficient forcefulness by the first-officer in correcting that self-deception
- This can be seen in pilot's lack of response to the co-pilot's prompts
- Pilot doesn't respond to co-pilot until the plane is already falling
- My reactions:
- To be honest, this article isn't very valuable
- Lots of speculation and going off on tangents
- Reading the NTSB Report indicates that while the pilot's decision to proceed with the takeoff in suboptimal conditions was the primary cause of the accident, there were other contributing factors:
- Improper de-icing procedures
- Improperly maintained de-icing equipment
- Excessive delay between pushback from gate and takeoff clearance
- Unstable aircraft design that was known to be vulnerable to ice-buildup on wingtips (indeed Boeing introduced a new anti-icing system for the 737 after this disaster)
- No way to determine that Pt2 probes were blocked by ice
- If you want a better example of crew-resource management issues, look at Korea Air 8509
- This is one of my objections to nonviolent communication. Sometimes a degree of violence is warranted to get the other person to acknowledge that you're disagreeing with them
- Other things to bring up at discussion:
- Three-Mile Island failure:
- Pressure-operated relief valve got stuck open
- Venting steam led to false water level readings
- Same inconsistencies in instruments; same confusion among crew, who persisted in trying to solve a problem that didn't exist
- Elites are elites for a reason
- It's possible that the people who are richer than you are just better, in every way
- This is so horrifying to contemplate that nobody talks about it
- It's not in the interest of those who are in the power elite to talk to you until you also make it into the power elite
- Eliezer is drawing conclusions off a biased sample
- He doesn't realize just how biased his sample is - the majority of any profession do not attend seminars and professional development conferences
- Finally, can we equate intelligence with competence, like Eliezer does?
- Moreover, intelligence just seems to make the inevitable screw-ups worse - look at LTCM, for example, or the financial crisis
- Self-made elites vs. elites who've inherited their position?
- How to produce a lot of social value for relatively little cost?
- It's possible to produce a lot value by expending lots of resources, but is there a way to get a lot of social value relatively efficiently?
- Add liquidity where needed
- Microloans or startup capital
- Isn't the impact of microloans somewhat controversial, at this point?
- Solve coordination problems
- Pool risks
- Insurance
- Hedges and contracts
- Provide information to allow others to allocate their resource better
- GlassDoor for companies
- Though, GlassDoor is a poor example, because it falls into the Yelp problem - the only people who comment are those who either had very positive or very negative experiences. Moreover, GlassDoor has an inherent conflict of interest, because it gets money by recruiting for businesses.
- GiveWell for charities
- Restructure choice sets so that our biases work for us instead of against us
- Richard Thaler won the Nobel Prize in economics for this
- Use status quo bias to encourage organ donation
- Use prize-linked savings accounts to encourage retirement savings
- Remove rent-seeking
- If it were that easy, it'd be done already
- Get rid of unnecessary occupational licensing
- Reduce transaction costs
- My reaction
- Overall, this is an okay article, but the "remove rent-seeking" bit was amusing
- More to the point, I think this misses the fact that you will face active opposition in doing all of these these things, from incumbents and regulators, and therefore the benefit of doing these things has to be equal to overcoming active opposition
- Being irrational is a good thing
- Rationality only makes sense in the context of a goal
- We don't have the time or capability to calculate probabilities and potential risks that come with every choice
- Even data based decision making doesn't inoculate against irrationality or prejudice
- Financial crisis
- Experts were convinced that it was statistically impossible for the events of the 2017 financial crisis to occur
- It's possible to convince yourself that you're being perfectly rational, when in fact you're overfitting to historical data
- Moreover rationality ignores the fact that decisions rest as much on subjective preferences as they do on objective facts
- People tend to rely on heuristics rather than statistics and this is a good thing
- Recognition heuristic is as good at predicting the winners of Wimbledon than ATP rankings
- Hyperbolic discounting is a good proxy for modeling uncertainties that may prevent us from getting a payout in the future
- Emotions are key to decision-making
- People who suffer damage to parts of the brain that are responsible for generating emotions find themselves unable to make decisions
- Emotions let people know what their preferences are, which is important for decision-making
- The marriage example is wrong. You would absolutely be a fool to go along with your emotions in that. People tend to be extremely good at fooling themselves when it comes to relationships, and often this self deception only wears away years (or even decades in) and leads to painful divorces
- Courage can be seen as excessive optimism, but it leads to great achievements
- Group selection vs. individual selection - just because it's good for the group to have some courageous members doesn't mean it's beneficial for each courageous individual to be courageous
- Decisions are made in a social context
- The social implications of a decision influence us as much as the actual consequences of the decision
- Our ability to agree easily with those around us is key to our ability to collaborate
- The Society of Friends (a.k.a Quakers) have come to the right conclusion on on a surprising number of things in a surprising number of domains
- Their virtues are those of liberalism, as are their vices
- Why should we respect the Society of Friends
- Proto-liberals
- Freedom of religion
- Freedom of thought
- Belief in individual liberty
- Belief that people had an "inner-light" that served as a connection to God and a moral compass
- Personal integrity as a radical practice
- Quakers had integrity to a fault
- Refused to sign letters with "normal" closing lines because they were considered to by insincere
- Refused to quote high prices at the start of bidding, because this was seen as a dishonest bargaining tactic
- Virtue over profit maximization
- Nonviolent social technology
- Quaker society relies on persuasion over force
- Quaker society treats dissent as signal to be processed rather than noise to be suppressed
- Humble Marketing
- Quakers don't market themselves - they focus on providing high quality goods and services and trust that people will find their way to them
- So, if Quakers are so cool, then why isn't Ben a Quaker?
- Saying, "I wrote about this argument elsewhere, so I won't repeat it here," is a dick move comparable with, "The proof of this lemma is trivial and left as an exercise for the reader."
- Quakerism is vulnerable to arbitrage
- If you allocate more resources towards good works, then someone else can make money by allocating resources towards works that oppose your good works
- If you volunteer to help people then you incentivize systems that rely on your volunteer efforts rather than systems that prevent the problems in the first place
- Vulnerable to specious guarantees - Starbucks can't deceive you about the quality of their coffee, but it's much more difficult to verify their "fair trade" claims
- Quakerism works well in a closed system (i.e. intentional communities or local production), but it tends to work less well in open systems where bad actors can exploit its vulnerabilities
- Quakers don't pay enough attention to the problems of ensuring the continued existence of their community - don't have children and are eventually displaced by those who don't share their values
- The problem that intentional communities have is that they're inherently going to conflict with the mainstream world, and that conflict will be an ongoing drain, unless they work towards economic autarky
- Has he actually read any histories of intentional communities? The only people who've actually pulled off the economic autarky thing are the Amish. Pretty much every other form of intentional community dies.
- Alternatives to Quaker values
- Puritans
- Don't exist don't any more
- Weren't very fun
- Very traditionalist - probably wouldn't be open to some of the more radical rationalist conclusions
- Jews
- Jews are drifting away from the practice of Judaism as they're forced to make compromises to remain compatible with a modern economy
- Haredi Jews are much more insular, but they're also much more dogmatic, and don't seem to produce as much material progress
- Academic communities - focus on the integrity of their intellectual production, but they're dependent on the outside world for economic inputs and new members
- Hippie communities seem to have the hang of living well, but don't seem to produce much in the way of material progress
- Burning Man
- Only lasts two weeks
- No artifacts of lasting value - leave no trace
