Fri 26 Oct 2018

Web3 notes: Nature 2.0

Josh Fairhead Public Seen by 251

These are notes from "Nature 2.0" node in web3 which is a rather interesting as a thought experiment. Especially along the lines of how do we design it so an automated planet acts in symbiosis with itself rather than exponentiate rent dissipation?


With the growth in AI we are seeing a movement towards smart vehicles. Its unlikely to be just smart Ubers; AI will also likely displace 13m truck drivers. Retail it will displace 4.6m jobs. It will also displace creative jobs and AI researchers too - lets start there.

There are a couple of ways to frame AI
* Replacing human cognitive behaviour (Turing test - can it fool a human)
* It can do “human only tasks” (Ariel design)
* It can do a task at a speed/accuracy/capacity not possible for a human (antenna design)
* As "a set of tools"
* Classification
* Linear classifiers
* Neural networks
* Deep learning
* Support Vector Machines
* Topology decision tree (Token Curated Registeries)
* Boosted Trees
* Random Forests
* Bagged Trees
* Regression
* Input variables for reverse search
* Knowledge extraction (e.g. XML harvesting)
* Optimisation
* Creative/Structural design
* As embodied agents (Artificial General Intelligence)

AI Sub Fields
* Machine Learning
* Neural networks / deep learning
* Evolutionary computation
* Swarm algos

DAO’s - Decentralised Autonomous Organisations (generalised)
* Blockchain automation
* Network of smart contracts

* AI on the blockchain
* Progressive iteration applied to value automation
* Two extremes:
* Numerous AI agents feeding a smart contract
* Numerous smart contracts singling AI agents
* Most likely a network like ant pheromone signalling (swarm algos)

Ways to arrive there:
* Start with a DAO and add AI
* Start with AI and add a DAO
* Start with SaaS, turn it into a DAO and add AI
* Start with a physical service like Uber and turn it into a DAO

To note:
* Organisational rights for DAO’s eventually become rights for AI
* Apply to nature and give the forest and rivers rights!

Nature 1.0:
* Trees
* Need nutrients
* Need rain
* Recreate themselves
* Are symbiotic
* “Non sentient"
* Self sustaining
* Anti-fragile
* Come in higher levels of integration
* Tree -> Forest -> Planet

Path of arrival to Nature 2.0 - Digital twinning of Biology and AI DAO’s
* Self driving cars - Smart
* Self owning cars - Smart with rights (new forms of personhood)
* New forms of personhood for self owning/regulating infrastructure
* Self owning forests, farms, rivers and common pool resources
* -> Higher levels of integration
* Profit? -> UBI chain?
* Basic income to higher levels of integration like leisure

Open questions and notes:
* Biological and mechanical symbiosis - what happens when the roads want to eat the forests?
* Can one hijack capitalisms short term mentality to create long term abundance through incentive models?
* Farm ICO -> buy farm -> add sensors -> add DAO -> input to AI model -> hires/fires/invests/purchases etc.
* Rivers in New Zealand now have rights via sensors
* Can we use cases like those to create jurisdictional arbitrage?
* Open motors and Robotic are focused on ride sharing cars.

* The presentation from Interlinked.ai is integrated into these notes also

Further thoughts on the above or the inevitable "redesigning nature"?


Josh Fairhead Tue 30 Oct 2018

Token perspective
• Store of value
• Security, asset, currency

Active Token /Access rights
• Usage token
• Work token
• Reputation token

Fungibility perspective
• Only quantity - can be merged/divided
• Non Fungible Token; can be distinguished as unique (ownership of object)

Legal perspective
• Is it money? Is it a security? Does it pass the howe test? Did people invest? etc.