I put some of my not very well thought out but ardently held ideas (100 year staking, exponential curves, Node Provider markets) into the machine and asked it to tear it to shreds. Some good stuff popped out…I don’t know what thread to put this in so it goes here. I think these are much simpler alternatives and more aligned with the incentives we want to give and community sentiment.
Proposal Sketch: Long-Term Alignment Without Governance Ossification
This is not a fully specified proposal, but a direction for discussion around improving long-term alignment in IC governance and node provider incentives, while preserving adaptability and decentralization.
The core goal is simple: reward credible long-term commitment without locking the network into irreversible assumptions.
It is also based on these assumptions:
- 2 weeks staking is a secruity nightmare…I’m shocked that it is even being considered.
- we WANT longterm commitment…why are we killing it?
- NPs got a sweet deal but its time to make it a market
1. Extend Maximum Staking Duration (With Exit Penalties)
Increase the maximum neuron dissolve delay significantly (e.g. well beyond today’s limits, potentially up to multiple decades).
However:
- Long-duration neurons must retain an early-exit option
- Early exits incur a time-dependent penalty (steep early, tapering over time)
- Penalties are burned or redirected to a neutral pool to avoid perverse incentives
This allows participants to express strong long-term conviction while preserving the system’s ability to adapt if assumptions change.
The objective is not permanent lock-in, but credible commitment with a painful cost to reversal.
2. Endogenous Long-Horizon Rewards Curve
Replace short-term optimized reward dynamics with a smoother curve that rewards patience and long-term alignment.
Key principles:
- Rewards increase meaningfully with longer dissolve delays
- No hard dependence on external benchmarks or fiat interest rates
- Parameters remain adjustable via governance, but change slowly
The intent is to reduce short-term cycling and governance gaming, while keeping issuance policy fully endogenous to the protocol.
3. Subnet Participation: Constraints First, Market Second
Subnet security and decentralization must remain non-negotiable constraints set by the NNS:
- Geographic diversity
- Jurisdictional independence
- Hardware and operational requirements
- Fault tolerance assumptions
Within those constraints, market mechanisms can be used to allocate limited slots.
Importantly:
- Markets should operate inside security boundaries, not define them
- Eligibility is determined by the NNS; allocation is competitive among eligible providers
This preserves adversarial assumptions while allowing price discovery where appropriate.
4. Node Provider Rewards: Baseline + Commitment-Weighted Component
Node provider rewards could be split into:
- A baseline component in liquid ICP sufficient to cover operating costs across approved regions
- A variable component that NPs tied to participation in active subnets and long-term commitments - **NPs bid on how much of this reward goes into a VERY LONG HORIZONED neuron.
Rather than raw utilization, variable rewards should emphasize:
- Reliability and availability
- Proven operational performance
- Willingness to accept longer-term reward vesting or delayed liquidity
This avoids incentivizing artificial load while still rewarding durable participation.
Why This Matters
If no participants — stakers or node providers — are willing to make long-term commitments even with reasonable exit options, that signals a deeper confidence problem worth addressing directly.
Conversely, mechanisms that make long-term alignment economically meaningful can shift behavior away from short-term extraction toward:
- Network reliability
- Tooling and ecosystem investment
- Reputation and institutional credibility
The aim is not to force commitment, but to make it rational.
Open Questions (Intentionally Unresolved)
- What is the right upper bound for meaningful long-term commitment?(I say 100 years signifies “FOREVER”…ICDevs would stake for 100 years…machine says that is dumb and impractical and suggested early exit fees that slacken as you get closer to your commitment)
- How steep should early-exit penalties be to deter churn without causing bank-run dynamics?
- Which commitment signals actually correlate with good node provider behavior?(Node providers should be developing software…I know a couple of them either directly or tangentially funded some software, but it hasn’t been clear, public, available, or broad)
- Where should markets be allowed to operate — and where must they be excluded?
This post is meant to frame those questions, not answer them definitively.
