1 of 18

Token engineering

Shruti Appiah Ethereum Waterloo ConsenSys

2 of 18

hey it’s Shruti

Education

Systems Design Engineering, UW 2018

Research

Complex systems <> Token mechanisms

Computational social science

Work

ConsenSys Labs etc.

Founded Ethereum Waterloo

Founded Prava

3 of 18

A crypto-economy is a complex system.

(-5,-5)

(-1,1)

(1,-1)

f(0,0)

Past does not predict the future

Dynamic, strategic behaviour

Connections matter more than the components

4 of 18

A crypto-economy is a complex system.

Non-linear & chaotic

Bottom up

Scale-free network

5 of 18

Why?

Analysis only gives 1 Nash equilibrium per month

Iterative crypto-economic design

Focus on building your system.

01

02

03

6 of 18

Can we anticipate all failures?

Collapse of Tacoma Narrows Bridge

7 of 18

How do I design a coordination system without enforcement?

The GOOD

:) Mechanism design!

:) Given: All my players/agents are rational and self-interested.

:) Objective: What’s the desired outcome?

:) Problem: How do I design a game that achieves said goal?

The BAD

:( Slow & hard

:( Fraught with assumptions

:( Life doesn’t have all the answers!

8 of 18

It’s an optimization problem.

9 of 18

Formulate your problem

? Domain

? Objective function

? Constraints

10 of 18

What algorithms suit?

Many metaheuristics exist...

Evolutionary algorithms

Genetic algorithm

Optimization

  • Ant-colony optimization

Good for pathfinding on graphs

  • Particle swarm optimization

Simulates social behavior - swarm intelligence also studied for AI

11 of 18

12 of 18

Be mindful of . . .

Overfitting

Solution: add more constraints

NP hard problems

Solution: Reformulate problem

If nothing works,

make your own token optimization algorithm.

13 of 18

Design Patterns

Curation

TCRs for list of good actors

Curation markets for determining popularity of an asset

Identity

Low level public keys

Personal identity & device identity

Governance

Aragon, EIPs, Colony

Proof of <???>

Data (human work) or computation (machine work)

.

14 of 18

Verification

General verification

Does your system work?

Blockchain-specific verification

Does your system work with Ethereum’s current state?

Scalability solutions are here to help!

  • Plasma
  • State channels
  • TrueBit

.

15 of 18

Validation

Does your system achieve what it’s supposed to?

Model your crypto-economy!

.

Agent-based models

Graph Theory

Systems modelling

16 of 18

Good luck.

Stay responsible.

17 of 18

Appendix

18 of 18

Blockcanvas

Identify the applicability of Blockchain technologies for your use cases.

Input the data from your own case and see how it fits.