Thread by @TaylorPearsonMe

Been thinking lately about frameworks that help explain the ideological divide in crypto.

One is a concept from engineering called the efficiency-robustness tradeoff. A thread. 👇

The Efficiency-Robustness Tradeoff (ERT), is a principle from engineering that given a constrained of resources individuals (and organisations and communities) must choose between being effective and being robust.

If you have 10 hours to write, you can make a decision between editing one article five times and your edits being extremely robust (unlikely to miss errors) or writing five articles one time each and being more efficient (more total output).

Neither of these decisions is “right” or “wrong.” The optimal point on the ERT spectrum is domain and outcome dependent.

If you are running a small consumer web app, having your server go down for a few hours every now and then is not that big of a deal.

It’s better to err a bit on the efficiency side of the spectrum because the costs of failure are low and spending extra resources isn’t worth it

If you a running a nuclear power plant or air defense system, the same point on the ERT would be a disaster.

You can’t have the nuclear reactors go down “every now and then.” Systems with high failure costs are extremely “inefficient” - trading off that efficiency for robustness.

Public blockchains are a new system and much of the disagreement seems to be about where on the ERT spectrum crypto networks should be.

One side (~Bitcoin maximalists) sees crypto primarily as a monetary technology and thinks that we should treat crypto networks more like nuclear power plants.

The relatively conservative development roadmap and focus on 2nd layers scaling solutions emanate from this belief in robustness at all costs. - "Better Safe than Sorry"

The other side (~Ethereum/smart contracts/web 3) sees crypto primarily as a computing technology, closer to a consumer web app.

They acknowledge that robustness is more important than on the consumer web, but thinks we should still allocate more towards efficiency than a nuclear power plant.

The more aggressive development roadmap and belief that scaling can safely be done at the base layer with minimal security/robustness tradeoffs all emanate from a belief that Bitcoin is "wasting" resources without achieving as much as it could.

The goal in both visions is to create the best possible outcome, which means that the system should be sufficiently robust, but not overly robust so as to “waste” resources.

The trick bit is that ,a priori, it is impossible to know how robust is overly robust.

The Lindy Effect is a good heuristic for non-perishable things like technology - the future life expectancy is proportional to their current age, so that every additional period of survival implies a longer remaining life expectancy.

If a book has been in print for 40 years, it is likely to stay in print for another 40.

If a cryptographic system hasn't been broken in 20 years, it likely won't be broken for 20 more.

For a technology that’s been around for a while, like bridges, engineers know pretty well at this point what the robustness/efficiency tradeoffs are.

For new technologies, it’s often that case that much of what appears efficient in the short term, is in fact just hiding risk somewhere and is really a ticking time bomb. See the history of Long Term Capital Management (Spoiler: it wasn’t very long term) - en.wikipedia.org/wiki/Long-Term…

I am more sympathetic to the notion that public blockchains are more like nuclear power plants than consumer web applications.

In general, I think modern society tradeoffs too much robustness to gain more short term efficiency because in many modern systems the gains are privatized while the losses are socialized.

If you know you are going to be bailed out, why wouldn't you take a bunch of risks in the name of efficiency over robustness?

I remain open-minded because I am more sympathetic to the robustness side of the spectrum in pretty much every domain so that may reflect more of an underlying ideology and view of the world, than a superior understanding of one specific domain.

H/t to @arjunblj and @yassineARK’s piece that took a different angle on the same question and stimulated a lot of my thinking - medium.com/@arjunblj/a-co…