From seeing like a state to learning like a machine.
A state can only make rectilinear decisions based on paper forms. Scalable but binary.
A machine can make curvilinear decisions based on digital data. Scalable and statistical.
Of course, some states have partially adapted & have some of their data online (albeit in often insecure databases).
But it’s just a retrofit. It’s not like the leaders are looking at metrics on the health and wealth of their constituents. Nor are they making decisions via code.
As software continues to eat the world, ultimately legislators themselves become engineers: the core developers of the digital platform that citizens run their lives on.
Changing a policy is then literally merging code. The gap between written law and citizen experience narrows.
Many legislative changes by legacy states have unintended consequences because courts disagree on original intent, citizens react in surprisimg ways, special interests game the system, etc.
But what if we took software seriously? Replaced the words of the legal code with code?
Precedents for this include the engineers updating the codebases behind massive global two-sided marketplaces like Uber and Airbnb, and blockchains like Bitcoin and Ethereum.
These economies are bigger than those of small nation states. And they are managed at root by code.
Pushing out an overnight change to 100% of citizens via federal law is like pushing untested code to 100% of users in production.
Many software methodologies can instead be adopted, such as testing new code/laws on opt-in cohorts and special innovation zones before scaling up.
In this model, the “president” is the founder of the blockchain-based digital platform you have opted into.
The “legislators” are the core developers who write the code.
And the “courts” are the nodes and miners that interpret the code.
Executive, legislative, judiciary.
It’ll be hard to retrofit these concepts onto all but the smallest states given the philosophical gap.
Easier to do it from scratch, in a parallel opt-in system, that proves itself online by attracting digital citizens till its internal economy is comparable to a legacy state.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
File sharing was once huge, and may become huge again.
The fundamental new primitive that blockchains offer for protocol design is tamper-resistant global shared state.
With this, someone could do a crypto Napster, KaZaa, or PirateBay. Probably pseudonymously, or in a country like Denmark with a "Pirate Party". Maybe not in music…
P2P, MVC, CBC
Blockchains allow us to combine the decentralization and programmability of the P2P era with the global state and monetizability of the MVC era.
For protocols with global state, the new architecture is then client-blockchain-client, or CBC.
The creator economy is becoming part of the cryptoeconomy. Because legacy social media platforms have no real concept of digital property rights.
Substack, Twitter Revue, and Facebook Bulletin are all great. But they are in a sense intermediate steps.
Ghost goes further. You can run it at a domain you control.
Decentralized social media platforms go further still. You hold the private keys.
Under communism, there was no such thing as personal property. Everything that transpired in the PRC & USSR was downstream of that economic illogic, the root cause.
The restoration of private property was one of the keys to unlocking the Chinese economy. npr.org/sections/money…
Longevity has the potential to be to traditional medicine what crypto is to traditional finance. It changes the terms of the debate.
Starting with Bitcoin's rejection of central banking & endless inflation, the cryptoeconomy has challenged virtually every premise of the state-controlled, paper-based financial system that we've inherited.
So far? Plenty of risks, plenty of loss — and undeniable progress.
The conventional macroeconomic wisdom is that high inflation is bad, but that deflation is also bad, so a little inflation is good.
But there's bad deflation, often due to contraction of money supply. And then there's good deflation, due to genuine productivity increases.
1) One day charts are highly volatile and we need to see where everything averages out. Some kind of drop looks real, but capacity may come online in other places.
2) The first chart (hashrate) is over years, the second is over months - otherwise you can't see the blip.
One somewhat vexing thing:
(a) when hashrate drops a lot, you want a quick difficulty adjustment
(b) but when hashrate drops a lot, blocks take longer to mine, so difficulty adjustment takes longer to come