I’m not sure whether your sequence will touch on this, but the things that make me hopeful in this space are not techniques and strategies for individuals (which might require training, or willpower, or shared values), but rather suggestions for novel institutions and mechanisms for coordination.
For instance, when you take a public goods problem (like pollution or something), expecting tons of people to agree or negotiate on how to resolve the problem seems utterly intractable, whereas if you could have started with a market design which internalizes such negative externalities, the problem might mostly resolve itself.
Since successful longstanding institutions (like nations) necessarily have a strong bias towards self-preservation, however, I don’t really see how most novel mechanisms could possibly be implemented (e.g. charter cities are a great idea, but pretty much all nations are deeply skeptical of them due to highly valuing their sovereignty).
One avenue that seems like it could have a bit more hope is in the cryptocurrency sphere, if only because it’s still quite new, plus it’s also inherently weird enough that people might not immediately balk at bizarre-sounding concepts like quadratic voting.
For instance, Vitalik Buterin’s blog contains many proposed novel coordination mechanisms:
Summary of the essay “Alternatives to selling at below-market-clearing prices for achieving fairness (or community sentiment, or fun)”: Why do people sell concert tickets below market-clearing prices? This has big negative consequences like incentivizing scalpers, but also some advantages like following some intuitive principles of fairness (e.g. not locking poor people out of the market), as well as more cynical reasons like “products selling out and having long lines creates a perception of popularity and prestige”; etc. So the post suggests a market design that allows selling at mostly market-clearing prices while still preserving e.g. fairness, and concludes with: “In all of these cases, the core of the solution is simple: if you want to be reliably fair to people, then your mechanism should have some input that explicitly measures people. Proof of personhood protocols do this (and if desired can be combined with zero knowledge proofs to ensure privacy). Ergo, we should take the efficiency benefits of market and auction-based pricing, and the egalitarian benefits of proof of personhood mechanics, and combine them together.”
The essay Moving beyond coin voting governance points out in an aside that cryptocurrencies invest ridiculous sums in network security (proof of work), e.g. here’s a chart of spending on proof of work vs. research & development. The difference is that network security was considered as a public good during design of the cryptocurrency protocols, while e.g. research wasn’t. So the former gets huge amounts of funding; the latter doesn’t. (Though the essay also points out that rewarding R&D explicitly would compromise their independence etc.)
Other mechanisms include quadratic voting, which was IIRC also used here on Less Wrong for the 2018 Review; as well as the related concept of quadratic funding.
To which extent these benefits will actually materialize is of course still an open question, but conceptually, this sounds like the right approach: align incentives, internalize externalities, consider public goods from the start, etc. Try to improve systems, rather than people.
Yeah this is largely what I meant when saying “I’ve updated towards ‘things like somewhat better markets.’” I have been reading Vitalik Buterin and agree that crypto is a pretty interesting testing ground for new coordination schemes. (another motivating example was microcovid.org, which helped streamline covid negotiation for large numbers of people)
Part of my goal of the sequence is to orient a bit on “how rationalists/EAs/longtermists might deliberately experiment with coordination schemes internally as they scale”, with part of the hope being that this enables them to perform better.
I think smaller-scale strategies are still important because (at current margins) many groups I know are small enough that scalable-mechanisms usually aren’t (immediately) the bottleneck.
I’m not sure whether your sequence will touch on this, but the things that make me hopeful in this space are not techniques and strategies for individuals (which might require training, or willpower, or shared values), but rather suggestions for novel institutions and mechanisms for coordination.
For instance, when you take a public goods problem (like pollution or something), expecting tons of people to agree or negotiate on how to resolve the problem seems utterly intractable, whereas if you could have started with a market design which internalizes such negative externalities, the problem might mostly resolve itself.
Since successful longstanding institutions (like nations) necessarily have a strong bias towards self-preservation, however, I don’t really see how most novel mechanisms could possibly be implemented (e.g. charter cities are a great idea, but pretty much all nations are deeply skeptical of them due to highly valuing their sovereignty).
One avenue that seems like it could have a bit more hope is in the cryptocurrency sphere, if only because it’s still quite new, plus it’s also inherently weird enough that people might not immediately balk at bizarre-sounding concepts like quadratic voting.
For instance, Vitalik Buterin’s blog contains many proposed novel coordination mechanisms:
Summary of the essay “Alternatives to selling at below-market-clearing prices for achieving fairness (or community sentiment, or fun)”: Why do people sell concert tickets below market-clearing prices? This has big negative consequences like incentivizing scalpers, but also some advantages like following some intuitive principles of fairness (e.g. not locking poor people out of the market), as well as more cynical reasons like “products selling out and having long lines creates a perception of popularity and prestige”; etc. So the post suggests a market design that allows selling at mostly market-clearing prices while still preserving e.g. fairness, and concludes with: “In all of these cases, the core of the solution is simple: if you want to be reliably fair to people, then your mechanism should have some input that explicitly measures people. Proof of personhood protocols do this (and if desired can be combined with zero knowledge proofs to ensure privacy). Ergo, we should take the efficiency benefits of market and auction-based pricing, and the egalitarian benefits of proof of personhood mechanics, and combine them together.”
The essay Moving beyond coin voting governance points out in an aside that cryptocurrencies invest ridiculous sums in network security (proof of work), e.g. here’s a chart of spending on proof of work vs. research & development. The difference is that network security was considered as a public good during design of the cryptocurrency protocols, while e.g. research wasn’t. So the former gets huge amounts of funding; the latter doesn’t. (Though the essay also points out that rewarding R&D explicitly would compromise their independence etc.)
Other mechanisms include quadratic voting, which was IIRC also used here on Less Wrong for the 2018 Review; as well as the related concept of quadratic funding.
To which extent these benefits will actually materialize is of course still an open question, but conceptually, this sounds like the right approach: align incentives, internalize externalities, consider public goods from the start, etc. Try to improve systems, rather than people.
Yeah this is largely what I meant when saying “I’ve updated towards ‘things like somewhat better markets.’” I have been reading Vitalik Buterin and agree that crypto is a pretty interesting testing ground for new coordination schemes. (another motivating example was microcovid.org, which helped streamline covid negotiation for large numbers of people)
Part of my goal of the sequence is to orient a bit on “how rationalists/EAs/longtermists might deliberately experiment with coordination schemes internally as they scale”, with part of the hope being that this enables them to perform better.
I think smaller-scale strategies are still important because (at current margins) many groups I know are small enough that scalable-mechanisms usually aren’t (immediately) the bottleneck.