Towards a Practice of Token Engineering

Towards a
Practice of
Token Engineering
Trent McConaghy
@trentmc0

1000x more data
The Unreasonable Effectiveness of Data
1000%
less
error!

Silo mo’ data
Mo’ accuracy
Mo’ $
Default incentive:
hoard the data

“Show me the incentive
and I will show you the outcome.”
-Charlie Munger

You can get people to do stuff
by rewarding them with tokens.
This is a superpower.

Change the
incentives!
Silo Pool mo’ data
Mo’ accuracy
Mo’ $

Early iterations:
Flailing
Can we
structure this
better?

Realization: Tokenized Ecosystems
Are a Lot Like Evolutionary Algorithms!
What Tokenized ecosystem Evolutionary Algorithm
Goals Block reward function
E.g. “Maximize hash rate”
Objective function
E.g. “Minimize error”
Measurement
& test
Proof
E.g. “Proof of Work”
Evaluate fitness
E.g. “Simulate circuit”
System agents Miners & token holders (humans)
In a network
Individuals (computer agents)
In a population
System clock Block reward interval Generation
Incentives &
Disincentives
You can’t control human,
Just reward: give tokens
And punish: slash stake
You can’t control individual,
Just reward: reproduce
And punish: kill

We can approach token design
as optimization design.

Steps in Optimization Design
1. Formulate the problem. Objectives,
constraints, design space.
2. Try an existing solver. If needed, try different
problem formulations or solvers.
3. Design new solver?

1. Formulation of an optimization problem
Objectives & constraints in a design space

2. Try an existing solver. Does it converge?

Example of a Successful Outcome

Token Design
as Optimization
Design

Steps in Token Design
1. Formulate the problem. Objectives,
constraints, design space.
2. Try an existing pattern. If needed, try different
formulations or solvers.
3. Design new pattern?

1. Formulate the Problem
(a) Ask
•Who are my potential stakeholders?
•And what do each of them want?
•What are possible attack vectors?
(b) Translate those into objectives and constraints.

2. Try Existing Patterns
1. Curation
2. Proofs of human or compute work
3. Identity
4. Reputation
5. Governance / software updates
6. Third-party arbitration
7. …

2.1 Patterns for Curation
•Binary membership: Token Curated Registry (TCR)
•Discrete-valued membership: Stake Machines
•Continuous-valued membership: Curation Markets
characterized by bonding curve
•Hierarchical membership: each label gets a TCR
•Work tied to membership: Proofed Curation Market
•Non-fungible tokens: Re-Fungible Tokens

2.2 Patterns for Proofs of Compute Work

Case Study:
Analysis of Bitcoin

Bitcoin objective function
Objective: Maximize security of network
• Where “security” = compute power
• Therefore, super expensive to roll back changes to the transaction log

Bitcoin objective function
Objective: Maximize security of network
• Where “security” = compute power
• Therefore, super expensive to roll back changes to the transaction log
E(Ri) α Hi * T
E() = expected
value
# tokens (BTC)
dispensed each
block
block
rewards
hash power of actor
= contribution to
“security”

Result of Bitcoin’s objective function:
People are maximizing security! = Maximizing electricity
More power than USA by mid 2019

1. Formulate the Problem:
(a) Who are stakeholders? What do they want?

Objective function: maximize supply of relevant data
Token rewards if: supply relevant data
Token rewards if: supply data, and curate it
1. Formulate the problem:
(b) Translate into objectives and constraints

Constraints = checklist:
• For priced data, is there incentive for supplying more? Referring?
• For priced data, good spam prevention?
• For free data, is there incentive for supplying more? Referring?
• For free data, good spam prevention?
• Does the token give higher marginal value to users of the network versus external
investors? Eg Does return on capital increase as stake increases?
• Are people incentivized to run keepers?
• Is it simple? Is onboarding low-friction?
(b) Translate into objectives & constraints

(b) Translate into objectives & constraints

(b) Translate into objectives & constraints – values too!

2. Try Existing Patterns
Some patterns:
1. Actor registry
2. Data registry
3. Actor registry + data registry
4. Data registry + free-as-in-beer data curation market.
Curation: Pay tokens to listen.

Key Question 1 2 3 4 5
For priced data: incentive for supplying more? Referring? ✖ ≈ ✔ ≈ ≈
For priced data: good spam prevention? ≈ ✔ ✔ ✔ ✔
For free data: incentive for supplying more? Referring? ✖ ≈ ✖ ✔ ✔
For free data: good spam prevention? ≈ ✔ ≈ ✔ ≈
Does token give higher marginal value to users of the
network, vs external investors? Eg Does return on capital
increase as stake increases?
✔ ✔ ✔ ✔ ✔
Are people incentivized to run keepers? ≈ ≈ ✔ ✔ ✔
It simple? Is onboarding low-friction? Where possible, do we
use incentives/crypto rather than legal recourse?
✔ ✔ ≈ ≈ ✔
2. Try existing patterns: evaluate on objectives &
constraints. None passed…

3. Try New Patterns
Some patterns:
1. Actor registry
2. Data registry
3. Actor registry + data registry
4. Data registry + free-as-in-beer data curation market. Curation:
Pay tokens to listen.
5. Data registry + free data curation market. Curation: Stake
tokens as belief in reputation. Auto CDN.
6. Actor registry + free&priced data curation market. Curation:
Stake tokens as belief in reputation. Auto CDN. “Proofed
Curation Market”

Key Question 1 2 3 4 5 6
For priced data: incentive for supplying more? Referring? ✖ ≈ ✔ ≈ ≈ ✔
For priced data: good spam prevention? ≈ ✔ ✔ ✔ ✔ ✔
For free data: incentive for supplying more? Referring? ✖ ≈ ✖ ✔ ✔ ✔
For free data: good spam prevention? ≈ ✔ ≈ ✔ ≈ ✔
Does token give higher marginal value to users of the
network, vs external investors? Eg Does return on capital
increase as stake increases?
✔ ✔ ✔ ✔ ✔ ✔
Are people incentivized to run keepers? ≈ ≈ ✔ ✔ ✔ ✔
It simple? Is onboarding low-friction? Where possible, do we
use incentives/crypto rather than legal recourse?
✔ ✔ ≈ ≈ ✔ ✔
3. Try new patterns: evaluate on objectives &
constraints

Objective: maximize supply of relevant data
• Reward curating data (staking on it) + making it available
• New pattern: Proofed Curation Market
E(Rij) α log10(Sij) * log10(Dj) * T *Ri
Expected
reward for user
i on dataset j
Dj = proofed popularity
= # times made dataset
available
Sij = predicted popularity
= user’s curation market
stake in dataset j
# tokens
during
interval

From AI data to AI services
Motivations:
• Privacy, so compute on-premise or decentralized
• Data is heavy, so compute on-premise
• Link in emerging decentralized AI compute
Objective function: Maximize supply of relevant services
=reward curating services + proving that it was delivered
E(Rij) α log10(Sij) * log10(Dj) * T *Ri
proofed popularity
of service
predicted popularity
of service

Design of Tokenized Ecosystems
From Mechanism Design to Token Engineering
Analysis: Synthesis:
Game theory Mechanism Design
Optimization Design
Practical
constraints

Design of Tokenized Ecosystems
From Mechanism Design to Token Engineering
Analysis: Synthesis:
Game theory Mechanism Design
Optimization Design
Practical
constraints
Engineering theory,
practice and tools
+ responsibility
Token Engineering for Analysis & Synthesis

Trent McConaghy
@trentmc0
• Token design ≈ optimization design
• So, approach token design as optimization design!
1. Formulate problem. Objectives, constraints.
2. Try existing patterns. Iterate.
3. If needed, try new design.
• This process helped a lot for designing Ocean (so far)
• Token Engineering = Theory + practice + tools + responsibility
Conclusion:
Towards a Practice of #TokenEngineering

Towards a Practice of Token Engineering

More Related Content

What's hot

Similar to Towards a Practice of Token Engineering

Recently uploaded

In this document

Towards a Practice of Token Engineering