Reading Notes
- The toxoplasma of rage does have one benefit - it makes people uncomfortable, which is one prerequisite for getting them to change
- Activist vs. passivist politics
- Scott takes a passivist approach to politics
- Repulsed by protests - doesn’t like the thought of people gathering specifically to demonize and vilify an other
- Others are too sensitive for activist politics - the notion that failing to take action to prevent a bad thing makes that bad thing your fault is psychologically triggering for many
- So do we deny the obligation to help?
- After all few of us impose a non-negligible marginal cost on the poor
- However, there does seem to be a deep-seated human desire to express gratitude and help the less fortunate
- Why do we privilege political action when we consider ways of improving the world?
- Compare attending a protest to donating to charity - charity is two or three orders of magnitude more efficient at improving peoples’ lives
- How much should we give?
- 10% is a reasonable Schelling point
- Donations don’t have to be entirely in terms of money - volunteer time works as well
- If everybody gave 10% of their income to charities, Moloch would die a swift death
- Moloch is the spirit of people responding to perverse incentives
- Giving to charity is an unincentivized action - anathema to Moloch
- Giving what you can individually to charity allows you to do good even when you find political action and mass movements to be uncomfortable or disturbing
- Separating charities and interventions allows charities to plug gaps in large scale interventions
- Example:
- AMF only purchases bed nets and does some regional distribution
- Relies on local partners to do individual distribution and need-assessment
- SCI doesn’t actually perform deworming treatment - only provides technical support and purchases medicine
- Separating charities from overall interventions allows charities to specialize and become more efficient
- We should evaluate charities not by their individual efficiency, but by the efficiency of the overall intervention that they are a part of
- Global catastrophic risks can permanently negatively affect the trajectory of civilization, even when they don’t lead to total extinction of humanity
- Why should we focus on catastrophic risks that may cause large numbers of deaths without causing humanity to die out?
- The probability of extinction-level events is small, compared to the probability of events that may cause hundreds of millions of deaths
- If a large number of people are wiped out, there is a possibility that civilization will not recover, or will recover in a way that stunts humanity’s long-term potential
- The combination of the above two factors means that threats that wipe out a significant fraction of humanity without causing extinction pose some of the same risks as extinction-level threats
- Basic framework and terms:
- 2 possible frameworks:
- Minimize expected deaths
- Maximize long term potential of humanity
- This article will be using the latter framework
- 2 levels of catastrophic risk
- Level 1: deaths of hundreds of millions or some billions
- Level 2: Complete extinction of humanity
- 2 schools of thought:
- Level-2 focus: efforts at minimizing global catastrophic risk should focus on level 2 events exclusively
- Dual focus: level 1 and level 2 events pose similar levels of threat to the long term development of humanity
- Why dual focus: there is some non-zero probability that civilization will not recover from a level 1 event
- Global catastrophes are more likely than extinctions
- Possible catastrophes:
- Pandemic
- Nuclear war
- Climate change
- Geoengineering
- AI
- In all of these scenarios (except maybe AI) a level 1 catastrophe seems more likely than full extinction
- General reasons to think that a global disruption might affect the distant future
- The world has had unusually positive civilizational progress over the last few hundred years
- There is little consensus about the mechanisms underlying civilizational progress
- Is the Industrial Revolution inevitable?
- There is essentially no precedent for a level 1 catastrophe
- The closest thing is the Black Death in Europe, and that occurred before the Industrial Revolution
- Specific mechanisms by which a catastrophe could affect the distant future
- Disruption of sustained scientific progress
- Scientific progress requires
- Scientists
- People willing to learn science
- People willing to turn scientific discoveries into inventions
- Institutional support for the above
- A disruption of any of the factors necessary for scientific progress could lead to the following failures:
- Risky stall:
- A stall in scientific progress makes humanity vulnerable to an extinction level catastrophic risk
- Resource depletion and environmental degradation
- Humanity runs out of some essential resource or degrades its environment too much due to insufficient scientific advancement
- Permanent stagnation
- Maybe we lose science forever and are just stuck at some particular level of development
- Disruption of sustained social progress
- Negative cultural trajectory - closed authoritarian societies gain some kind of permanent edge over open societies
- Would probably lead to a reduced rate of scientific progress
- Increased inter-state violence
- Irrevocable technological mistakes
- Weaponized AI
- Biological weapons
- Potential offsetting factors
- Could a level 1 catastrophe make humanity more resilient against level 2 catastrophes?
- Probably not
- Reactions to level-1 catastrophes are probably over-specific to that catastrophe
- Overall, industrialization has been a net positive for humanity, even taking into account environmental degradation
- A level 1 catastrophe seems as likely to set back preparations for dealing with critical technological junctures as it is to allow more time for preparations
- Conclusions and strategic implications
- Events with hundreds of millions dead are much more likely than extinction events
- These events could derail civilizational progress
- There is essentially no precedent for events of this scale
- Therefore a dual approach to X-Risk is warranted
- Cosmopolitanism is a moral stance that gives the interests of other nationalities a weight equal to people of one’s own nationality
- Pretty much unanimous support for cosmopolitanism among EAs
- However, most other people do not believe in cosmopolitanism
- “Anti-foreign/anti-immigrant” doesn’t carry nearly the same level of stigma as “racist”
- Analyzing policy with a cosmopolitan lens can be enlightening
- Allows EAs to find allies in places that they may not expect
- What does cosmopolitanism say about policy?
- Countries should give more moral weight to the citizens of other countries, especially when waging war
- Cosmopolitan civilian test for proportionality in war - would civilian casualties be considered proportionate if the civilians were of a different nationality?
- EA should be more open and explicit about cosmopolitanism