Chainflow’s Take on Replicated Security

Update (March 6th, 2023): After confirming that more than 2/3 of active Cosmos Hub validators must choose to validate new consumers chains for these chains to go live, we updated our vote on Prop 187 to "Abstain" in order to reflect that, while we still have open concerns about the specific implementation described in the proposal, we support the general direction of moving to Replicated Security.

Replicated Security is Exciting, Transformative, and Flawed

Chainflow joined the Cosmos community in 2017. Over the years, we've been fortunate to witness—first-hand, as a Cosmos Hub validator (find us, here)—the evolution of the Hub network and the broader Cosmos community.

In a recent post, we talked about our journey growing up in the Cosmos, and our excitement for one of the biggest steps to-date in its evolution: Interchain Security (ICS). In short: ICS introduces new ways for Cosmos chains to share security features, like validator sets.

Since we published that earlier post, the first ICS feature to be introduced has been officially renamed to "Replicated Security." And now, with the introduction of Cosmos Hub governance proposal 187, the Cosmos community is finally able to decide how Replicated Security will make its way to the Interchain, after months of discussion and development.

We see the benefits of Replicated Security, and recognize that it can become a key competitive advantage for the Cosmos Hub. We also appreciate all the hard work that's been done through the years to get us this far.

From the beginning, recruiting a competent validator set has required significant resources from networks wanting to launch in the Cosmos. ICS greatly reduces this barrier to entry.

Yet, the adoption of Replicated Security raises issues we feel need to be addressed before moving forward. We're concerned that Replicated Security, as we currently understand it, will:

  1. Reduce opportunities for new validator operators, concentrating power in the existing Cosmos Hub validator set
  2. Raise operational costs for validator operators, shrinking already-thin margins across the board
  3. Eliminate choice for Cosmos Hub validator operators, forcing them to support new chains that may not align with their values

At the core of these issues is the requirement that all Cosmos Hub validators must validate on every Replicated Security "consumer" chain, without the ability to opt-out on an individual chain basis.

Given this requirement, we won't be able to vote affirmatively for Replicated Security adoption—at least in its current state. The rest of this post explains why.

More Chains, More Validators

During the earliest Cosmos testnets, the primary focus for validator operators was on becoming a Hub validator. But, Zaki reminded us from time to time that operating a validator on the Hub wouldn't be the only opportunity for infrastructure operators.

Over the years, the number of non-Hub chains has grown considerably. Since each new chain has needed to build its own validator set, many validators who weren't able to enter the Hub's fixed validator set (currently limited to just 175 active validators) have been able to validate on non-Hub chains, and gain a foothold in the Cosmos ecosystem.

This has introduced the community to many well-deserving validators that contribute significant value to the chains they operate on. Many, if not most, of these up-and-coming validator operators may have never been discovered had operating on the Hub been the only way to support the Cosmos.

Replicated Security will make it such that new Cosmos chains no longer need to build their own, new validator sets; instead, they can become "consumers" of the security offered by the Cosmos Hub's validator set. As a result, the responsibilities of—and the power yielded by—the Hub's validators will grow, likely further entrenching the power dynamics already in place.

As our ongoing tracking from shows, stake doesn't move around much on the Hub (or, for that matter, other leading Proof-of-Stake chains). That means we're not likely to see many new opportunities for up-and-coming validator operators to break into the Hub's active set.

In other words: For up-and-coming validator operators, Replicated Security may result in fewer new validator sets to join, and fewer opportunities to break into the Cosmos Hub's validator set.

As vocal advocates of the need to #KeepStakeDecentralized, this power shift gives us pause when considering whether to support Replicated Security in its current state.

More Chains, More Resources, More Money

Replicated Security, as constructed in Prop 187, requires Cosmos Hub validators to validate on every "consumer" chain that launches. Naturally, this requires the validator operators to spend more CAPEX and OPEX resources on supporting these new chains.

The more networks an operator supports, the more servers, tooling, and human attention they need to manage their infrastructure well. These requirements translate, simply, into higher CAPEX and OPEX costs. Yet, many validators already operate at a loss, at least in these early days.

It's likely that Replicated Security will encourage more new Cosmos chains to launch. This, in turn, makes it likely that the associated percentage of successful chains will decrease.

All these factors combine to increase the obstacles that validators face to becoming profitable, or even breaking even. This is particularly concerning for smaller, independent validators, who have operated historically on thin margins.

What's not obvious to many Cosmos participants, though, is that these independent validators are critical to decentralizing networks. They typically contribute an outsized amount of value back to the chains they support—especially when compared to exchanges, staking-as-a-service providers, and large validator companies.

Losing smaller validators would result in the loss of invaluable contributors, while also delivering a severe blow to the decentralization of the Cosmos.

No Values, More Problems

At Chainflow, we put our principles before profit. In our values, we focus heavily on the core tenets of the crypto revolution that we're here to build the foundation for. In short: We're here to help build a better future—one that's more equitable, inclusive, and fair than the current economy, and the power structure it enables and upholds.

As crypto matures—and particularly as the corporate "Web3" narrative takes hold—the percentage of people starting projects that hold similar values seems to be decreasing. The "Web3" gold rush is on. This brings all types of speculators and others who are driven first and foremost by cashing out—all following the typical VC, Silicon Valley playbook.

One important way that we reflect our values is by very carefully selecting the networks we support. We don't work with projects whose values don't feel like they have the potential to align with ours over the long term. This approach has, time and time again, helped us avoid a number of the critical and high-profile pitfalls that crypto has experienced over the years.

While we hope that each consumer chain that launches is values-aligned, the Hub's governance history, combined with the continued "Web3" shift toward a money-first focus, leads us to believe that we'll likely not align with every "consumer" chain that launches. Having to validate on every consumer chain presents a significant risk to our values-aligned selection process.

Voting No, At Least for Now

Ultimately, we're voting "No" on Prop 187 due to this risk to our values.

We believe that allowing validators to opt out of indivdual "consumer" chains is a critical second filter that the Cosmos needs. For example: If a consumer chain passes a community governance vote, yet can't get enough validators to support it, maybe it's a sign that the chain shouldn't launch.

While voting "Yes" for the current version of Replicated Security would feel in conflict with our values, we feel, nonetheless, optimistic about the Cosmos's future and will continue to support the Hub as a validator, regardless of Prop 187's outcome.

If Replicated Security is implemented in its current form, we will use that as an opportunity to advocate strongly for chains we feel values aligned with, while raising red flags for those we don't. If it isn't implemented in its current form, we look forward to working with the community to refine the approach to maximize the benefit this powerful new feature can provide to the Cosmos Hub, and the broader Cosmos ecosystem.

We've been supporting the ecosystem since its earliest days, and we'll continue to support it, and the Cosmos Hub, regardless of the outcome of Prop 187. However the vote works out, we'll remain active, engaged, and vocal.

If your values align with ours, please consider supporting our work by delegating your ATOMs to Chainflow's validator, here.